Playing Bots Instead of Humans to Improve at Chess is the Best Way

I've come to this opposite conclusion from the consensus view.  Hear me out.  So basically, what I learned in the chess dojo training club I took part in on and off over the past few years is that long games with deep thinking and deep calculation is the best way to grow, particularly if you then also pair that with long post game analysis and extensive annotations of that game.  If you can also have a chess coach review it with you and go through it move by move giving their additional feedback, then you are really maxing out chess growth with these approaches.  Yes, chess books and puzzles etc are great too.  But the biggest growth is playing long games and analyzing them deeply.  This trains you to think deeply about the moves and not just guess or go with intuition and rush through things.  It also helps you uncover where your bad tendencies are and weaknesses are which helps you work on those.  It also helps you remember the games and problems for the lessons they bring since you are spending much more time on it.  

So all that being said, it turns out that long games are really hard to find online and are inconvenient pretty much necessarily as it pertains to human opponents.  You have to either schedule them with somebody - so have to work with them to figure out a mutually workable time, or you have to go to a scheduled OTB event or you have to potentially sit in long queues.  Even if you do queue 60 minute games (which is at least decently long), 90+30 is considered much better for growth.  And that option isn't even available on standard queues on chess.com so you'd have to manually find someone to join that format and can't just random queue it especially at higher ratings where the player pool gets slimmer.  If finding the games is inconvenient or you have small children and don't get out much, you simply won't be doing much of these ideal and vital long games - thereby hindering your growth.  

So all that being said, in comes the bots option!  With bots, you can play as long as you want and pair with them at any time day or night.  They are always available for you at a moments notice - unlike human alternatives.  This is a HUGE upside and asset because it means you will play WAY more games in that long format than would be possible with humans.  Next, with humans, if you do a long game time control, up to half the whole game's duration is waiting on THEM to make their moves.  It's not just you that gets your desired longer time, but the opponent also then can take forever to move.  This can be downright annoying or boring to wait on the opponent to make their move and so this would be more likely to tempt you to multitask or w/e.  I mean you could think about the position and think about future moves and stuff while you wait though which is probably best, but still, that is one of the common downsides of longer time controls is waiting for the opponent to move.  Well with bots, that is a TOTAL NON-ISSUE!  You could take 3-10 minutes on your move, and the bot is not getting bored or side-tracked while waiting - which is nice and now you don't feel bad for taking so long like you might with a human opponent - and when you finally make your move, the bot does its move INSTANTLY!  This means you have ZERO downtime between your move and you beginning your next move selection.  This means that the longer format games are MAXIMUM engagement and MAXIMUM efficiency for your learning process, since you don't have to wait at all between moves.  This then builds your mental endurance since you go from thinking hard on a move to thinking hard on the next move with no breaks in between.  That's a nice boost you wouldn't really have in a game vs a human - unless you thought hard during opponent's turn but generally I think most people don't do this but instead take breaks or even go walk around in OTB situations while it is opponent's turn in longer games.  

So if you view long time control as best, as I have been convinced it is, and you do NOT lack self control and discipline to stick to that best practice type of game - refusing to get sucked into addictive blitz and bullet formats - as I refuse to, then bots really are better on so many levels than human opponent alternatives.

Now for potential downsides - which really are not downsides at all.  One could say: "well if you just play bots for a long time, how will you track your rating!?"  That's a good point.  If you just play bots exclusively for a long period of time, you really don't have an accurate way of gauging your rating/player strength.  That is a downside in a sense.  But that is also a HUGE upside in another sense.  I read just the other day recently someone said if only they could disable ratings in their chess.com account they would do so because obsessing over their rating was ruining their experience.  They literally sought to hide it.  I compare this to people wanting to hide their viewer count when they livestream.  It can be demoralizing to have few viewers when livestreaming and feel like you are talking to a small or non-existent audience.  So people hide their viewcount and just play act like they have a massive audience so that they can perform at their best as a livestreamer.  In the same way, some people don't want to see their rating.  Now granted, without seeing it, it is harder to know how much you are improving - but I must ask: can't you see the improvement just by observing your growth in understanding?  Can't you see it by your increasing pattern recognition speed and breadth?  Can't you see it from your increasingly deep calculations?  Can't you see it from the fact you are beating bots more consistently and with higher accuracy and fewer blunders?  Can't you see it as you begin to beat higher ranked bots?  The answer is YES!  You can see the improvement and gauge it in those ways.  But wouldn't you be curious to know where you REALLY stack up after a while in terms of the human rating pool of all players?  Isn't that fun to see that go up too?  Of course.  The curiosity on that would grow more and more because inevitably, you know you are growing and going to see your visible rating go through the roof when you come back to human games again.  And you can't wait to see it!  But this growing curiosity and excitement to see your visible rating go up BECOMES a great carrot on the end of the stick and a fun thing to think on and speculate about.  That becomes a mini enjoyment in itself!

So when DO you find out your visible rating?  Well to start, note that I'm calling it visible rating instead of rating.  When you play against bots or train outside of rated games, particularly for a long season of time, your playing strength will grow massively over time and so your visible rating stops reflecting your actual playing strength after a while.  This gap grows and grows.  You can only speculate on how much it has grown - which as I mentioned is a fun mini enjoyment in itself.  But the fact is, as this gap grows, your visible rating stops being as relevant to you since it is not up to date after a while.  So then we need a way to differentiate that visible rating from your true player strength.  Hence, the idea of calling the two ratings the visible rating and the hidden rating.  The hidden rating is the rating you would be if you were to go back to playing human games and played enough games to reach opponents of your new level.  The visible rating is the rating you left off at before you went into bot training mode and left the live player pool play.  

As an aside to the supposed downsides we are addressing, I wanted to bring up this one upside:  human games vs human opponents are more stressful generally.  At least for most people who aren't playing as frequently and are serious about training and improvement and have big goals for their rating.  Also, the tighter time controls add to this stress.  But playing the bots with time controls off just lets you play at your own pace and focus on really mastering the game, not distracted by constantly looking at the clock and time management.  That side of the game may make it fun in certain ways, but if you are really just focused on mastering the ideas and positional understanding and strategy and finding tactics and just becoming a phenomenal player, the constant distractions of time management is just distracting you from the game itself.  The time management game is like a mini other game in itself but is just taking away your focus and hindering your learning process.  Sure, when being tested in a tourney or prepping for some type of tourney play or w/e, it can be good to reintroduce it.  But if you are just interested in growing as a player of chess in general and becoming stronger overall, clock management should just be kept OUT IMO.  So playing the bots with no time control is superior then.  And really, you still probably will end up playing about the same speed as you would were time control there anyways.  But the key thing though is you are no longer DISTRACTED and constantly a little stressed out by that ever clicking tick tick in the back of your mind which leads to so many - "well I guess this is just good enough" and moving the piece since you feel the pressure of that tick tick in the back of your head and this is hindering you from really finding the best move you can and really pressing yourself to master the game at that moment, really moving forward with more confidence on every move, not just settling for a best guess given the time taken type of deal all the time.  This will be very beneficial for you as a player in the long run.

The next supposed argument against bot training is they allegedly play like Magnus every single move and then randomly do a SUPER obvious blunder, hanging a queen, then go back to playing like Magnus.  So according to this objectively false claim that keeps being repeated, every bot plays engine best moves every move of the game and then does a free piece giveaway then goes back to engine best moves again.  I know for a fact this is false.  In the massive number of bot games I've played, this has proven completely untrue.  In fact, now that I'm analyzing my games extensively when doing my annotations post-game, I have found that the opponent bot does not do engine best moves very often at all.  I'm playing mostly 1300 level bots and they are playing "good" moves, inaccuracies, excellent moves, and best moves with a human-like mix of all of them.  And the occasional blunder maybe once a game does match a human of this level I think too.  So it really is giving you a mixture and certainly is NOT just playing like Magnus then a random obvious mistake then Magnus again as is being falsely claimed!  It is just objectively not true.  These bots really are just playing average moves alot of the time and so I am sick of seeing this false depiction that keeps being repeated about playing bots.  

Next: I want to address the idea I hear all the time about the need to lose games to stronger opponents to grow and improve.  It's NOT TRUE.  Hear me out:  every move you make in chess is a mini game in itself - like the first move of a tactic puzzle - it is a challenge in itself.  And so if you view a game of chess as a series of mini games where each move is a game in itself, then even if facing a weaker opponent, you played many games you could have lost or won since you played many moves that were either good or bad or somewhere in between.  That said, I try to find the BEST move or at least top 2 or top 3 moves every move in the game.  So even if I play a opponent that is my strength or weaker, and even if I beat them fairly easily, that does not mean I can't have a challenge or learn since every move was a challenge to find the BEST move for that move.  So if you are really trying to min/max your every move, and play with the highest possible accuracy every game, then the strength of the opponent's moves whether really good or really bad, does not really dictate your challenge of finding the best move.  Well if they are so tough on you that the game gets so complex that finding the best move gets a bit too hard for you in that situation, that would be an exception.  But the point is, focusing on the weak opponent example: if your opponent is weak and you are able to dictate the game to some degree due to your superior strength, that does NOT mean you can't grow or be challenged since you STILL have the opportunity to PERFECT your board awareness, deep calculations, tactical mastery, positional understanding, pawn structure, piece activity maximization, move order, intermezzo move planning, prophylaxis moves, forming batteries, and on and on it goes.  All the concepts you'd read in books and want to gradually integrate so much stuff into your games, a weaker opponent that does not have your back against the ropes pounding you the whole time will enable you to explore and implement these various ideas in a relaxed environment so that you manage to get them more and more to all fit into your gameplan!  When queing rated games, you get equal or stronger alot of times and the games are usually quite hard.  But when playing bots, you can purposely play bots of lower strength and master so many concepts this way.  And you CAN LEARN MASSIVELY while doing so.

Next: is it fun to get the snot beat out of you every other game compared to just being on a dominating winning streak?  Sorry but I prefer the latter.  As a former athlete, I can assure you that winning seasons where we dominated every opponent was far more fun than losing seasons.  The victory celebrations and whatnot were epic whereas the slamming of lockers and cussing and misery after losses was not fun at all.  Not to mention the crying and loss of hope and loss of confidence.  So then, for someone who wants to focus on game mastery and implementation of so many concepts you learn in chess books and lessons and coaching and whatnot, there is SO MUCH to remember and try to implement into your game... is it constructive to be getting beat on and in fight or flight mode most of the game barely able to remember the fundamentals due to your panic?  NO.  That is not constructive.  It is not optimal nor ideal.  You will NOT learn as much.  Also, seeing your rating go down after a demoralizing loss is NOT beneficial or encouraging or uplifting and is a MAJOR distraction from your learning process.  It is also not FUN.  If you are not having fun, then you won't play as much and won't enjoy the process as much as you could have.  This will hinder growth.  So losing is not good for growth.  Winning is best for growth.  Disclaimer:  losing SOME is good for learning how to handle loss and not be a sore loser etc.  But my above statements assume you have already mastered those life lessons and are not a sore loser etc and are mature in that area already.  In that case, losing would not be beneficial at least to the degree that winning would be.  CONFIDENCE is a great asset in such a psychological game and the less losses you take the higher the confidence you'll have and the better advantage and enjoyment you will enjoy.

So how can you pretty much just be a predominantly winning chess player?  Great question!  Here is my proposal:  you play bots primarily for long seasons of growth paired with coaching, long games + long analysis and annotations after and books and puzzles and lessons and w/e else.  And you just get stronger and stronger.  But all the while these games are verse bots significantly weaker than your player strength is now.  And you win most games if not every game.  You can set them to be somewhat challenging, but at a level where you still dominate every game for the most part.  This means you keep fun HIGH and confidence PEAKED and hype high.  During this time you can even delude yourself into thinking you are like national master level by now or w/e.  You can inflate your sense of progression to unrealistic highs and let your chess ego grow massively Dunning-Kreuger style.  And it can be FUN to do this!  Ride this high!  Take it to the bank!  You get to have that winning season everyone loves in sports.  And as we already proved, you get to MASSIVELY learn the ENTIRE TIME due to the min/maxing of every move approach and becoming more and more dominant over the opponent with more and more best moves and brilliant moves and rarely good moves and less excellent moves over time since you are finding straight up BEST moves more and more, proving your understanding of the game is vastly growing.  And without the distractions of time management, losing, tilting, hope dashing, curb stomping of your ego and rating, etc, you can just improve and improve unhindered without all the pain and hard knocks in the way of idealized and efficient growth.

So then the game plan becomes this:  beat the heck out of bots for a long season - grow a good 200-300 points in hidden rating, then come out of bot training ready to dominate the humans at your visible rating level and dominate them.  Once your visible rating gradually increases to your hidden rating level, and losses start creeping in too much, pull back, cut off human play, and go back to bot grinding again to grow another 200-300 points in hidden rating > then go play the humans again and dominate them again.  Rinse and repeat.  This way you always stay a step ahead of the competition and always dominate both humans and bots.  This maximizes your confidence and chess ego and ensures maximum fun and zero tilting.  I suggest having a 300-400 rating gap always between your hidden rating and your visible rating - if not more.  This way when you do come out of bot training mode and into human play, with your 300-400 rating gap established, you can stomp your human opposition with relative ease taking very few losses relatively speaking.  The aim would be to have like a 80% win ratio I think.  You'd be more like a prize boxer in your win loss ratio then which is SUPER fun and you'd feel like just a beast at chess doing this!  And when the rating gap closes down to like 200 or w/e, and the losses start creeping in and your rating growth slows, that's your queue to shut it down and go back into bot training mode to re-establish your 300-400 point gap between secret rating and visible rating again so you can come out and dominate again.  

Next, under this system, you can dominate the chess scene and do so primarily with online play - which is ideal since it's more efficient and cost effective so you can play more.  And then when you reach title level, you can go to OTB tournaments and just dominate and get your title then come home victorious with title in hand.  Then grind up to your next title benchmarks and go dominate tournaments again to get that title.  Rinse repeat.  This way the tournament scene also is dominated the same way you dominated the bots and dominated the human opponents you played online.  Everywhere you go and everything you do is just DOMINATING!  That is FUN to me.  

Okay so back to more common misconceptions about bots... people are saying playing bots is too different than human opponents so it doesn't translate.  I call BS.  Sure it might be a bit different, but to say that it won't translate is just short sighted.  If you play bots and train in the ways I've laid out, and your hidden rating goes up 200-300 points and you get to 300-400 points buffer between your hidden rating and visible rating and go play human opponents 300-400 rating points behind your hidden rating, it doesn't matter if they play a bit differently than the bots you are used to, you are 300-400 points above them in rating!  Their tactics and ideas being different you will be able to see right through because you are so dominatingly higher than them in your player strength overall that you will laugh at their attempts to thwart you!  So while the bots may have a bit different approach, that still will not matter when it comes to playing humans as long as you are ensuring your hidden rating is sufficiently higher than the humans to make the transition downsides trivial and meaningless due to the skill gap involved.  So that solves that.  Plus, I think that the play style of bots is growing more and more human-like and the fact that top human players study bot lines so much and whatnot makes them more often than not the same as a bot if you think about it.  They become middle men for the bots whose lines taught the top player what to do and so when the top players are playing you are getting the moves of a bot at the end of the day that were memorized in advance.  So as you go up the ranks, creativity and non-botlike play is not so much an issue to overcome as that only is noticed mostly in certain parts of the middle-game after memorization has ended and we are "out of theory" mode.  So this notion that getting overly used to playing bots failing to prepare you to play humans and thwarting or hindering you long term is nonsense.  You will be FINE!

I want to now point out a really cool additional bot/AI help I found:  when analyzing my games and annotating them, I first annotate everything I think about the move and the position and whatnot and my plans at that point in the game etc.  I then check the engine to see if it had better move ideas and I note those and discuss any of those that stood out to me and write down my questions about that move or my explanations for why it was superior if I understand it fully.  All of this is in my annotations.  THEN I go to chatgpt and feed the board position and all moves of the game into it's prompt and ask it to deepdive that move for me.  It then gives a 1 page paper writeup of the move, the board position, the strengths and weaknesses for both sides, comments on perhaps superior move alternatives, etc.  I have a text to speech reader read its response to me while I look at the position and visualize everything it is reading to me as my AI coach.  Then once that is done, I copy my annotation for that move into chatgpt and let it generate a 1 page response to that.  I use a text to speech browser extension for chrome to read that out loud for me too while I look at the board and see everything it says.  Doing this has been an amazing way to get extra insight and training and a verbal explanation to go along with engine recommendations which would otherwise be not explained at all.  Now chatgpt is not perfect and will say alot of stuff that makes no sense or get who is white or black mixed up or w/e, but I ignore all of that or fix it in my mind to sort out what it meant to say as I listen and still get the grasp or gist of where it is going with this.  I also can say that probably 70% to 80% of what it is saying is true and the other 20-30% of rubbish can be fixed up in your mind if you say "oh it meant to say black there" or w/e and this way it can get up to 90% accurate with your mental edits of what it's saying.  In any case, to have a coach that tells you 90% accurate stuff with great insights is AWESOME and to have a coach that is available FOR FREE 24/7 any time you want to go over every game you play for HOURS at no cost to you.. and 90% accurate things to say... that is INSANE VALUE.  So I have been loving this.  My games are probably 1.5 hours right now maybe 1hr sometimes?  Maybe 45 min sometimes?  But I try not to go much lower than that.  If you go much lower than that, you aren't taking your time and trying to deeply understand everything and deeply weigh all possibilities and responses and do deep calculations and analysis.  If you don't put in this mental work you are being mentally lazy and not going to grow as much.  You have to put in that mental sweat work to grow more.  So yeah... lets say 1 hour games then on average and then 2 hours of analysis and annotation.  This is probably minimum you should do to maximize growth.  GM Jesse Kraai does like 4 hour games and 20 hours of analysis per game.  He goes way deeper than someone of my level.  But that is the extreme version of what I'm doing and the chess dojo teaches to do with games.  That is getting big time mental work in and deep understanding can start to creep in that way.  Not just spamming games and turning brain off.  Although that can work too for some people I guess but is ghetto and not recommended by the best coaches generally.

Which brings me to my next point:  people say if you play chess bots you won't grow or learn because you will not care enough to try your best and will not take your time and will just mindlessly play.  I say NONSENSE!  That is YOUR CHOICE if you approach bot play that way.  Your own laziness and lack of ambition, discipline, and motivation to grow.  I do NOT do this.  Now that I have been taught to take my time and think deeply and not pick the first good move I see but to think about the enemy response ideas and counter-play risks and overal position and overal piece activity and longer term planning and on and on it goes, now that I know how much more can be thought about for every move, I no longer find it responsible or a good use of time to just guess and move or w/e.  I want to MASTER the game and try to MASTER every single move.  My goal is for every move to be perfectly accurate and have super high accuracy games and if the engine did not agree with my move when I check that at the end, I want to figure out why and fight to understand why and learn in so doing and often I AM able to figure out why the engine moves ranked 1, 2, and 3 in the analysis window were better than mine and why and what made the top ranked move better than the 2nd ranked move and so on.  I seek to understand that during analysis and am more and more able to figure it out too!  How are you gonna say THAT is not growth?  LOL!  Can't grow from bot games?  GIVE ME A BREAK WHAT A JOKE TO SAY THAT.  So yeah, you can treat bot games as more serious than tournament games if you wanted.  There is NO LIMIT to how seriously you can take bot games and moreover, there is no limit to how seriously you can treat EACH AND EVERY MOVE you make in a bot game.  Not just that game overall but EVERY SINGLE MOVE in that game can be treated very seriously and your very best effort put in.  And doing it this way, you can really get your move accuracy numbers to skyrocket!  And THEN you will start to play like the big boys of chess.  So just because you play a bot weaker than you does NOT mean you can't challenge yourself to play every single move as solidly as possible.  And when you analyze and see better moves you could have made over and over in the game than the one you chose and you see why the move you failed to spot was better and how impactful it would have been and understand why, then you will evaluate your flaws.  

Which brings me to my next point:  people think that getting beat is the only way to see flaws in their game.  Like the only way to know if you are playing flawed moves is to get checkmated by the opponent.  That is RIDICULOUS and completely false.  You can see all the flaws in your moves in a game you dominated the opponent with little challenge and checkmated them and the opponent was a lot weaker than you by simply bringing that game into analysis window and analyzing it deeply.  Every single time you did not pick the best move you failed to find the best move.  You did not pass the chess puzzle for that move.  You get a big red x for that move.  Even if the move you picked was decent, it could have been better.  I mean it should be at least top 2 or 3 of the engine recommended typically.  Unless it is a rare case where the engine is bugging with some nonsense 16 move checkmate you should have spotted LOL.  Of course then we have an exception.  But overwhelmingly the majority of the time the top couple engine moves you should have seen and made and you can be taught that way or punished that way without the crushing blow of being checkmated!  You don't have to tilt or lose rating or anything to learn guys!  You just have to properly analyze your games to learn and see your flaws.  Your flaws do NOT have to be punished by way of a game loss.  It is enough to be punished by viewing your imperfections during post game analysis!  And all that said, the very notion a weaker player overall will NOT punish flaws in your game is also flawed.  I still have the weaker bot punish me in games when I did not see something coming or w/e.  I still have the evaluation bar swing to a less winning evaluation when this happens or even to a losing evaluation.  So you still get the feedback and punishment for mistakes.  But the key difference is, you still win the game in the end if the opponent is weaker.  So you get the best of both worlds.  You still get mistakes punished and you still get to win and have max fun by dominating and winning all the time.  This overly simplified notion that a player has to be stronger to teach you a lesson is NONSENSE.  And besides, you could purposely setup losing positions and then begin bot play if you really wanted to practice comebacks.  But I think they will come organically at times as long as you occasionally screw up big or have a bad game you'll get your chance to try to come back.  It will just happen but just not as often as if you played competition stronger than you are.  And that's fine.  You don't need that situation to happen on a regular basis (that painful, grueling playing from deep behind position)!  It will happen anyways whether you seek it out or not.  For example, when people you face online cheat in a game by using an engine - which is something notoriously happening in the chess world.  In those cases, even though you tried to queue with the higher hidden rating than your opponent and your own visible rating, you will still be now facing a much harder challenger since he's cheating.  So you'll still get your occasional mega challenge of a strong player then.  Or if you go to a tournament and someone 600-800 points higher rated than your visible rating plays you in a open section or w/e.  Now you get your higher rated player challenge - and since you have 300-400 points higher hidden rating, you still have a decent chance even then.  Or if some young chess prodigy who is severely underrated is paired with you on their climb up to their hidden rating, boom - you get your chance then.  So it's going to happen regardless.  I'm just AVOIDING IT as it is not fun to have to get spanked ruthlessly and humbled in a game!  Let's avoid that and have it be the exception to the rule rather than a normal thing!

Another common misnomer is "every bot is just blundering so obviously".  Sure, perhaps at the first levels there are obvious blunders, however, intermediate bots are not necessarily blundering so obviously in my experience.  And the human players blunders are sometimes more obvious than the bot ones.  So really the whole blundering bot thing is being vastly over exaggerated and someone's first impression of some low level bot they are automatically assuming applies to just all levels and that's just not true.  It's a bad stereotype that is not applicable much of the time.  And as I said before, the bot play is not 3600 at all times then hard coded blunders then back to 3600 play again.  That's just absolutely false.  They are playing normal good moves and ranked 2-6 moves.  Not just best moves constantly etc as is being falsely claimed.  

One big problem is noobs come in and see the early blunder bots and complain all bots are that way.  Then top players never play bots so they still have the same memory as the noobs from when they were noobs and played bots so they are equally biased to the same false conclusions.  Also there is the fact that bots are NOT the same as they used to be 10 years ago, they are getting better and better at playing more naturally and believable and human-like over time due to better coding.  This means a lot of the hate they get is based on obsolete, non-existent bot AI that no longer exists.  And yet people who have stopped playing against bots are perpetuating these no longer applicable hot takes on bots and since they are old school experienced chess players, everyone just mindlessly believes them despite them speaking out of ignorance and not having much bot playing experience to bring to bear anything relevant or up to date to say about it.  So all of these factors are spoiling the pool of opinions being presented whenever threads come up with titles like "is playing bots a good way to improve at chess".  And this blog post aims to stomp out all of these falsities and bring to light the truth on these matters.  Which is that playing bots almost exclusively in LONG games with DEEP post game analysis and study is the BEST way to grow and most efficient way to grow and most relaxing way to grow and most enjoyable way to grow and fastest way to grow and funnest way to grow and cheapest way to grow.  Just better in every category.  Yet almost NOBODY is doing it.  

But I am changing that.  I am doing it.  So is another guy named HighCue on chess.com.  He's doing it.  And I want to see more examples of serious chess improvers with big goals and aspirations to be titled players doing it.  Lets prove this false narrative wrong.  The chess world almost unanimously HATES the bots but they are all wrong about them.  Bots RULE for chess improvers!

Edit: one more thing:  another MASSIVE upside for long games with a bot is that for a person with a busy schedule but still insisting on very long game formats, you can literally just hit pause, go do other things with your day, and come back since there is no clock.  And this isn't rude since unlike with a human opponent, the bot has nothing else to do and did not schedule an uninterrupted game with you.  So you don't have to block out half a day for your long format game but can break up the game into segments that work around your schedule more!  And still maintain the long format game style and approach despite this.  So this makes long format games able to fit into a normal busy life and no longer has to wait for a full day off on a weekend or a full evening blocked out and dedicated to it.  And because the long game can then be done on a normal busy day by breaking it up some, you once again get to do them MORE OFTEN which means more chess improvement and more growth!  A HUGE win!  And as mentioned earlier, since the time the bot spends on moves is zero, you cut your long game in half in total time that has to be put into it and the only half remaining is the time YOU take for YOUR turns in the game.  So you have WAY more control of the total duration of the game and it is ALOT shorter than a human vs human long game where you also have that same amount of time to move personally.  Probably HALF as long to play it then.  This means that for the same time commitment in game play, you can DOUBLE the total games you play in this long format which means you can DOUBLE your efficiency and improvement by way of long games goes way up and is better in every way!  Heck, the second half of the game that just got erased since your opponent moves instantly, using that saved time, you can do most of your post game study/annotations during that time so that in the time you would normally have just played a long game you are now doing the long game and the long analysis and annotations, thereby greatly improving efficiency!

One more thing:  I want to really emphasize that even if you do queue 60 minute games on chess.com - the longest default time duration option available where you might get a pairing in a decent wait time - even if you do go this route as I've done several times, it's STILL tough to really get the feeling of being unrushed and thoroughly thinking through and understanding all factors you can at all times within reason.  Why?  Because even though you have an hour in total, one: you have in the back of your mind that 90+30 exists for a reason and if you use your time freely, really going deep into it, you really could run out if you don't watch it - hence the reason 90+30 exists and is a big deal in OTB high level play.  Second, and more importantly, in the back of your mind you know that if you spend alot of time on a move, your opponent IS going to be annoyed 9 times out of 10.  They will regret queuing the longer duration game and even wonder if you went AFK.  They cannot see you after all.  They might wonder if they are wasting their time!  And this DOES place a added stress and pressure on you since you do want to consider their feelings in the matter.  Yes, you do know you have the right to use your full hour, but you do also know that using large chunks of your time on a given move can and most likely will annoy your opponent and take away their fun and doing so then becomes a mixed bag.  You may feel the need to think more at some critical moments but feel pressured after a while to just move something so as to not anger your opponent.  And the longer you take to decide the more pressure you feel since the more time has expired.  So the time taken and pressure grow together and the pressure to move growing more and more at a key moment in a game is a MAJOR distraction that is lowering the odds of you making the best possible move at this key moment in the game and thereby ruining your experience in many ways.  You usually WILL just give in and move something last minute and when you do, since you just rushed last minute and winged it, sometimes you didn't even take the time to make sure they could not respond with something devastating to your position.  Your hasty last minute move, done out of consideration for your opponent's time and frustration levels just lost you the game since you just majorly blundered.  This has happened to me repeatedly!  So then I'm not even playing at anywhere NEAR my full potential at that point and I tilt the rest of the game and/or resign early.  Now my rating is tanking against a player I am significantly stronger than but just screwed up due to the pressure and time conflicts.  Yes, in a vacuum, you can tell yourself you have every right to use up your 60 minutes as you see fit.  And you can force yourself to ignore your opponent's sure frustration.  However, it feels a bit selfish and its hard to do it in practice, even if it is required that you do this in order to play to your true potential and get results that reflect your true strength.  This constant tug of war and major annoyance to online queueing long game time controls is a massive hindrance to playing humans IMO.  The only practical way around it is to schedule in advance a 90+30 with someone who really wants a OTB tournament type of game with you and really emphatically agrees that you can take as much time as you need and they want your best and you want their best and time is not a issue as you have both blocked out the whole half day for this or w/e and plan a extensive post-mortem session with them and everything up front.  Only then would I feel totally free to really use up my clock with no guilt at all.  But doing that is a massive commitment and much harder to find such a partner.  So once again, the bot playing option REALLY shines as all of these issues are instantly solved with the bot!

One more thing:  at the very most beginner level of chess playing, using the bots to learn with suggested moves and comments on and takebacks on is a GREAT way to learn and grow without getting beat up on in the human player pool and you get to do so while winning every game due to all the help and takebacks.  You are "cheating" with the help and takebacks, but you don't have to suffer all the emotional hurts of losses this way.  And this goes along with my theme of winning most games and just running on a non-stop dominating streak your whole chess career with my approach above.  BUT, once you start to feel confident on the basics and enter advanced beginner or intermediate level, eventually I recommend turning off all help and playing games through to completion and getting back real feedback this way on your total accuracy and move quality during your post game analysis phase.  With this accurate feedback, I think you will gradually begin to be able to estimate your hidden rating with at least some rough ballpark of accuracy.  Your puzzle rating, game accuracy, game win/loss ratio, level of bots you are beating consistently, etc will all form little clues that together can help you estimate your hidden player rating.  It can be done strategically with all of these clues I think.  And this will be necessary to do with some degree of ballpark accuracy if you want to be able to maintain the 300-400 rating point buffer mentioned earlier.

One more thing: on the topic of the alleged importance of getting beat in chess often so you can get punished for poor move selection and that negative reinforcement teaching you to stop blundering or w/e... I already proved that's not true.  But moreso, if you take finding best moves seriously as I do, if you take move accuracy percentage seriously as I do, then every single move during analysis that wasn't top engine move or at least top 3 engine moves, every move that a game review called "good" or "inaccuracy" or God forbid "mistake" and WHAT HORROR: "blunder", heck even "excellent" is less than best... if you take these critiques seriously, then each and every one of them is a punishment.  So you get punished massively in the analysis phase for every less than perfect move, anything not called "book" or "best" is a punishment.  So the RIDICULOUS false claim that you have to LOSE THE WHOLE GAME in order to be punished for bad moves I have now completely refuted and needs to die.  It is just copium anyways guys - lets face it.  People that lose a lot of games and hate losing invented a way to pretend losses are great by claiming they are the only way to truly grow as a player.  COPE MORE PLEASE!  *Facepalm*.  Just win most of your games and take your beatdown in the analysis about how your win could have been MORE PERFECTLY EXECUTED.  If you don't see less than perfect execution as a punishment, then your standards are so low, no wonder you don't grow much in this game.  This is a game that demands darn near perfection to go far.  So stop considering a won match as a win if the execution was not off the charts for your level.  You can get all the punishment you want by viewing move ratings this way.  And all the while still taste the sweet taste of victory if you follow my program laid out above!  And that victory in the final outcome of the game is the spoon full of sugar that lets the "good move, inaccuracy move, mistake move" etc punishments/medicine go down.