PoB 9/15: 4. Fantasy Chess

I don't even understand the question.
Do you mean like a fantasy football league, only chess? Or do you mean some kind of fantasy chess variant? Or do you mean like "who would win if Capablanca played Fischer?"


The bad thing about fantasy chess is that the best players aren't strong relative to the weaker players, since they don't play on the same boards. The only stats that chess players accumulate are wins, losses, draws... So what would make having a 2600 rated player better than having a 2450 rated player?
It would be better in a lot of cases to have the best 2nd board player in the league than to have the 2nd best first board player in the league, since you will end up with more wins. Does that make sense?
I'll be interested to see David's design. Please mention it on the blunders show so we don't miss it.

This sounds like fun. I remember Liz Vicary talking about participating in something like this.
Can non-premium members play? If so, I've got dibs on Amanda Mateer.

yeah, i'm envisioning any chess.com member could participate. although you'd be divided into ~16 member leagues :) you could also compare your stats with fantasy teams in other leagues.

Just a few thoughts (a warning, I didn't watch the show so I'm not sure what options were discussed). Assuming you're using the USCL there's probably only 3 ways to do a fantasy league.
1) Pick-em league. This would work just like NFL or NCAA pick-ems, with the option for a 'spread' for each match. Each matchup is listed, each participant picks which team they believe will win. Straightforward.
2) Pool pick-em (I'm not sure what the proper name is). I've never seen this version on any of the usual sports sites, but a number of newspapers run them. You put each chess player into a 'pool' of them with similar skills/boards (IE you could put the top 4 first boards into pool 1, first boards 5-8 into pool 2, etc.) and each participant selects one player from each pool. This is usually the one that rewards knowledge, since option 3 can tend to devolve into a slight bit of luck and/or punish people for being in a bad position in regards to others with knowledge of the sport. This is usually the one you would use when comparing a large number of players.
3a) Snake-draft (or Auction draft, but that would be extremely difficult to set up). Most common (iirc) fantasy football version. 16 people draft 1-16, then alternate order. There could be an option that each player would have to start x number of non-GMs akin to positions.
The main problem with fantasy chess choosing players as opposed to football players is that there's a lot more ways to score points in football. For instance, take a running back: Points for rushing TD, Recieving TD, running yards, recieving yards, per catch (sometimes). I'm not sure how easy it would be to have similar categories for chess players or to even figure them out (pawns taken, exchanges won, miniatures).
If I were to guess, I'd have to say that option 2 would probably be 1) the easiest to plan out, and 2) the most applicable to chess while still maintaining some feel of choosing players instead of teams,
not excatly like football or others, but i know that people play fantasy gold by picking 3 players and you total the scores of each of those three to see who wins. for instance if you chose golfers A, B, and C who scored 71, 72, and 73, then your total score would be 216. unlike other sports though, everyone can pick tiger woods if they want too.
not sure if this would work in chess.

Yeah obviously it would be better to have a higher rated player, but in these things, teams are capped. So say you get $100,000 to spend on how-every many players, the better ones are more expensive, but generally perform better.
Perhaps the actual chess players earn points of good/interesting moves, and lose some for mistakes? +Win/loss/draw.
Please invite me David! =]
Yes, the problem with doing some kind of chess fantasy league is that there are so few statistics - you can go by money winnings, points scored, performance rating, Rybka matchup rate ;) (to be clear, that's a joke), and... that's about it. Different stats will reward different kinds of players, though if you're trying to make it at all balanced, probably always people will look for underrated players to make up their rosters. Also, when does a season start and end?
Of course, the CLO (chess life online, the USCF website) or someone related thereto has hosted fantasy chess for the US Champs each of the past few years, but they haven't stuck to a solid format - nothing's worked all that great.
I think that what you want to have is some kind of free-for-all system, where everyone is in one big group. Whoever runs the whole thing will have to pick which events that the game applies to. The player picks whichever players they want from the event (probably you have to pick a certain number - say four), and then at the end, you total up the points they've scored and calculate two things: 1)the performance rating of your team (i.e. the average of the perf ratings of each of your team members, weighted by the number of games they played) and 2)the expected performance (or true rating) of your team (i.e. the average of the ratings of each of your team members going into the event, weighted by the number of games they played). Whoever has most outperformed expectations wins. You may even get to have a little more of a fantasy team field if you let players somehow choose to play certain players in certain matchups or on certain days, but in this case you probably also want to have some way to ensure that you aren't just picking to "start" your team in only games they have white, or something to that effect.
Topic 4 from yesterday's Pardon our Blunders chess.com/tv show:
What would make a good fantasy chess competition? (if anything)