Puzzle rush difficulty curve and the strikes system

Sort:
Brinerot

First of all, how come there are tactics puzzles rated as low as 100 elo on the site? A few of them were harder than the ones rated 2000+ I've seen on tactics trainer.

Second, the difficulty is relatively easy with mates-in-one, hanging pieces, recaptures, obvious knight forks and so on and at some point skyrockets to counterintuitive sacrifices, complicated sequences involving calculation of long lines (4-5 moves+). It happens in an instant and there's nothing in between.

Finally, the three strikes ruin my runs when I'm at my A game, while time gets the best of me before long when I'm doing badly. It's the same problem with tactics trainer, the clock dares you to make a move which results in more mistakes than usual. I know what you're saying, "don't move until you see it", but there are some problems where the situation is so counter-intuitive you don't really have a choice. Did I mention you can't really see the solution when you get something wrong, so if it's some pattern you don't know, you don't even get to learn it?

My proposed change? A time penalty of 15-30 (+-, don't know what a good value could be) seconds per mistake. The time one wastes to arrive at a wrong solution, losing the point, is also a penalty so things should be fine. I understand the purpose of strikes, to make it so people can't just breeze through with the first checks and captures they see, hoping luck will be on their side, but the current system is way too stressful, let alone having only 3 strikes.

 

notmtwain

They are the same problems you see on Tactics Trainer. Please post an example of a killer 100-rated problem. 

Your Puzzle Rush score says your highest level solved is around 1200.  In tactics trainer, you would solve those easily. It is only the time factor which makes them seem difficult.

Brinerot

notmtwain: Trust me, I've had quite a few instances where sub-1000 problems were as hard or even harder than the average 1800-2100 problem I see around on TT.

 

DeirdreSkye: A number of different problems are necessary, I definitely agree with that. I was talking about the strikes system and the difficulty curve anyway. And yeah, if by improving you mean reaching IM+ FIDE, then sure I do need some damn serious study until that.

notmtwain
Brinerot wrote:

notmtwain: Trust me, I've had quite a few instances where sub-1000 problems were as hard or even harder than the average 1800-2100 problem I see around on TT.

 

DeirdreSkye: A number of different problems are necessary, I definitely agree with that. I was talking about the strikes system and the difficulty curve anyway. And yeah, if by improving you mean reaching IM+ FIDE, then sure I do need some damn serious study until that.

I would just like to see a killer 100 point rated problem.  Remind yourself to post one here the next time you see one.

Brinerot

DeirdreSkye: I suppose after IM if you're ambitious enough you can reach GM. Sorry to disappoint you, but there are several intermediate levels between mine and IM. At best, I'm an intermediate player, below advanced, below expert, below master, below IM.

What I meant with "I'd need serious study to reach IM or more" is that improving as a weaker player is much easier than going for example from early GM days to the cutting edge of people like Magnus or Aronian for example.

Amador1955

Personally, I have no problem with how puzzle rush is structured. I'll point out that if a tactic baffles me I sometimes just enter a move at random to prevent the loss of time. So the three strikes and out seems a good way to prevent me from doing that more than twice.

JamesColeman

I've done Puzzle Rush 1126 times and I've never seen a 100 rated position that was anything more than capturing an obvious piece or a very obvious mate in 1. I also like it the way it is.

stiggling
Brinerot wrote:

First of all, how come there are tactics puzzles rated as low as 100 elo on the site? A few of them were harder than the ones rated 2000+ I've seen on tactics trainer. Some puzzles are artificially hard because they're mined by an engine, not a human. I blame chess.com for not finding a way to filter these out over the years.

 

Second, the difficulty is relatively easy with mates-in-one, hanging pieces, recaptures, obvious knight forks and so on and at some point skyrockets to counterintuitive sacrifices, complicated sequences involving calculation of long lines (4-5 moves+). It happens in an instant and there's nothing in between.

In other words at some point it crosses the threshold of your ability. Same thing happens for all of us, just at different levels. Even if some of the puzzles are crap, the ratings they get from years of tactics trainer usage tend to be spot on, so once they pass a certain point you're basically screwed. This is why it's very hard to increase your PR score.

 

Finally, the three strikes ruin my runs when I'm at my A game, while time gets the best of me before long when I'm doing badly. It's the same problem with tactics trainer, the clock dares you to make a move which results in more mistakes than usual.

Well, it's called puzzle rush for a reason. You're supposed to rush. It's not to make you better at chess it's just for fun.

 

I know what you're saying, "don't move until you see it", but there are some problems where the situation is so counter-intuitive you don't really have a choice.

"Don't move until you see it" is bad advice for puzzle rush. What you should do (or what I do at least) is calculate about 2-3 moves deep. If it still seems like a tactical-ish solution then I just go for it. Once the position is on the board it's easier to see the ending sequence.

 

Did I mention you can't really see the solution when you get something wrong, so if it's some pattern you don't know, you don't even get to learn it?

You can easily review the puzzles you missed by clicking on the red icons.

 

My proposed change? A time penalty of 15-30 (+-, don't know what a good value could be) seconds per mistake. The time one wastes to arrive at a wrong solution, losing the point, is also a penalty so things should be fine. I understand the purpose of strikes, to make it so people can't just breeze through with the first checks and captures they see, hoping luck will be on their side

What? That's not the reason at all. And if you're not guessing on the first few puzzles then you're just bad, IMO. Puzzles rated as low as 500 don't have complicated solutions. Essentially just find a way to give check then play it. If you're unlucky enough to fail then you can always try again next session.

 

but the current system is way too stressful, let alone having only 3 strikes.

Ok, so for you personally it's stressful, but that's not a good reason to change it for everyone.

For most people stress = drama = fun.

 

Brinerot

stiggling: Props for a well-structured response, not so much for the dickish, passive-aggressive attitude towards me. I won't even take the time for a proper reply for you, I wish I had time to waste though.

stiggling

shrug*

gingerninja2003

Puzzle rush wasn't designed to be easy (or hard). It was designed to be fun. 

Acroduster

Maybe they should have a millennial version where there is no strikes and every move you make is a winner? Would that make you happy 

stiggling
Acroduster wrote:

Maybe they should have a millennial version where there is no strikes and every move you make is a winner? Would that make you happy 

Every time you play it unlocks a new achievement or you get a trophy.

Every puzzle you fail, it doesn't tell you, instead it ends in success every time, and the next puzzle is easier.

Scores only go up, never down. Even if you solve 10 puzzles, if your personal best is 14, then it will say you solved 15.

The leader board only shows people who have scored lower than you, and when you visit someone's profile if their PR score is higher it's invisible.