I'm confused about the daily puzzles

Sort:
mikmaxschess
It seems like the Daily Puzzles are problematic, in that a lot of them require you to figure out the best string of moves... But also the AI often makes suboptimal moves. This was really, really apparent on the August 1st puzzle (Todays,) where - After your first move - the AI blunders the position from being 6 points down (which is bad) to Mate in 1. (Which is obviously worse.) I suck at chess, but I was able to spot a better move for the computer in about three seconds - And running it through an engine agrees with me. ("Better" in this case was easy because anything that's not a forced mate sequence is better.) This is kind of frustrating since I am usually only able to figure out a couple moves ahead, (Like I said, I'm not good at chess,) and so when I'm trying to figure out a puzzle, or even looking at the solution after I've given up or solved it, I can't tell if I really played the "best", and I can't figure out what would have been better.
mikmaxschess

Uh... Wow. Ok. That was kind of rude. I'm not attacking the developers, I'm just saying there's sometimes an element to the puzzles that I find fruatrating and don't understand the point of.

KeSetoKaiba

I actually enjoy the daily puzzles, and I especially liked the puzzle you are referring to (No need to worry 8/1/2018). Yes, I found the correct "solution" in one try (fairly quickly) as well. However, keep in mind that "real" games seldom follow "best" moves all the time. In fact, I prefer the mate in one finish to the puzzle. The purpose of daily puzzles is to teach a theme or idea/concept. 

In that particular puzzle (spoiler alert follows), the Queen couldn't be taken because mate in one follows. Therefore, the "idea" taught in this puzzle is that the Queen was immune to capture. By creating the puzzle to play a "worse" move (capturing the Queen), the puzzle is essentially asking you if you calculated the finish (mate in one) rather than moving blindly. If the puzzle move (better) was decided to go down by 6 points of material, then the one solving the puzzle may not have even noticed that the Queen was immune to capture here; I like chess.com 's decision in this puzzle to ensure the one who solves this saw the follow-up from taking the Queen. The goal of the puzzle creator should be to ensure the desired theme to teach was "solved" and taught; the goal is understanding the nature of the position - not necessarily playing the "best" move if the opponent does the same because not all opponents will play with precision. As a chess player, we need to learn how to capitalize on less precise responses - and this puzzle does this.

Preggo_Basashi
mikmaxschess wrote:
It seems like the Daily Puzzles are problematic, in that a lot of them require you to figure out the best string of moves... But also the AI often makes suboptimal moves. This was really, really apparent on the August 1st puzzle (Todays,) where - After your first move - the AI blunders the position from being 6 points down (which is bad) to Mate in 1. (Which is obviously worse.) I suck at chess, but I was able to spot a better move for the computer in about three seconds - And running it through an engine agrees with me. ("Better" in this case was easy because anything that's not a forced mate sequence is better.) This is kind of frustrating since I am usually only able to figure out a couple moves ahead, (Like I said, I'm not good at chess,) and so when I'm trying to figure out a puzzle, or even looking at the solution after I've given up or solved it, I can't tell if I really played the "best", and I can't figure out what would have been better.

Because

1) they can only show one line
2) the mate idea was the key pattern to solving the puzzle
3) if it showed the best play:

 

 

Then you'd have a horde of beginners saying

"The daily puzzle is so dumb, black could just capture white's queen"
"This solution makes no sense, white won a bishop, but that's not much, the game goes on"
"Obviously moving away from the bishop is stupid, these puzzles are so stupid I don't get it"

etc.

Preggo_Basashi
mikmaxschess wrote:
so when I'm trying to figure out a puzzle, or even looking at the solution after I've given up or solved it, I can't tell if I really played the "best", and I can't figure out what would have been better.

The moves it accepts for your side will be best. There's no need to worry about that.

 

As for the opposition's moves, in every line you'll achieve an easily winning position, so for the main line of the puzzle they often choose to show the main idea or theme. Sometimes this means the opponent's moves aren't best, but in any case it's a winning position for you. You can try to work out the sidelines on your own as part of the exercise (this is always useful).

 

At least, this will be the case when they're doing it right. That's not only chess.com, but quality puzzles you find in books for example.

mikmaxschess

Thanks, everyone, for the responses. I think I'm getting what you're saying - As long as the puzzler (me) solves the best line, it shouldn't matter what the computer does.

SeniorPatzer
mikmaxschess wrote:

Thanks, everyone, for the responses. I think I'm getting what you're saying - As long as the puzzler (me) solves the best line, it shouldn't matter what the computer does.

 

Well, it is good to see/know how to punish all the variations that the computer might have chosen too.

BryanCFB
Postafi wrote:

When you're confused about daily puzzles, just imagine yourself having to come up with a new or used puzzle every day (including Sundays) acceptable to every puzzle fanatic.

 

If Chess.com needs any help coming up with better daily puzzles there is a forum:  

Daily Puzzle Submissions! *Be Involved*

currently with 128 pages and over 2500 posts.  Has Chess.com ever used one of these posted puzzles for their daily puzzles?  I'm not sure but I imagine there are plenty of worthy puzzles there.

chessmastersgod11

So if the puzzle that sticks is wrong why do u have to lose 15 to 18 points? Then only get 5 when I get it?  I went from getting good puzzles to 20 bad puzzles in a row losing over a100 points.. I prefer learning with my own games

johnworldesen2200
Preggo_Basashi wrote:
mikmaxschess wrote:
It seems like the Daily Puzzles are problematic, in that a lot of them require you to figure out the best string of moves... But also the AI often makes suboptimal moves. This was really, really apparent on the August 1st puzzle (Todays,) where - After your first move - the AI blunders the position from being 6 points down (which is bad) to Mate in 1. (Which is obviously worse.) I suck at chess, but I was able to spot a better move for the computer in about three seconds - And running it through an engine agrees with me. ("Better" in this case was easy because anything that's not a forced mate sequence is better.) This is kind of frustrating since I am usually only able to figure out a couple moves ahead, (Like I said, I'm not good at chess,) and so when I'm trying to figure out a puzzle, or even looking at the solution after I've given up or solved it, I can't tell if I really played the "best", and I can't figure out what would have been better.

Because

1) they can only show one line
2) the mate idea was the key pattern to solving the puzzle
3) if it showed the best play:

 

 

Then you'd have a horde of beginners saying

"The daily puzzle is so dumb, black could just capture white's queen"
"This solution makes no sense, white won a bishop, but that's not much, the game goes on"
"Obviously moving away from the bishop is stupid, these puzzles are so stupid I don't get it"

etc.

 


If it takes the queen Bg5 will be checkmate