Draw problem? Incentivize wins!

Discussion of anything and everything relating to chess playing software and machines.

Moderators: Harvey Williamson, bob, hgm

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
Posts: 99
Joined: Sun Apr 26, 2020 9:40 pm
Full name: Brian D. Smith

Draw problem? Incentivize wins!

Post by Cornfed » Sat Oct 17, 2020 6:54 pm

We have all heard the options to 'change the game itself'- from 'no-castle' to '360' and in-between for top level tournaments. Whatever their merit, these actually change the game we have come to know and love. These ideas seem popular among:

1. Those who tend to not play competitively/seriously anymore.
2. Weaker players who…may find the traditional game too difficult to play as well as they would like.
3. People looking to popularize the game (even if it means changing it).

Those of us who have become fairly competent and have invested so much into the game naturally tend toward keeping the game as it is.

I think you look at different top level tournaments over the last few years you may conclude that if there is a ‘draw problem’, it is not due to the game itself... rather, it is your basic tournament 'structure' that contributes to this.

If you look at high level closed tournaments – particularly those with many participants - we (and the participants) know at some point if they are likely to finish towards the top or if they are destined to 'just collected another paycheck'. This knowledge can negatively influence the course of games toward, if not ‘draw’ at least toward ‘boring chess’ – particularly for those likely destined toward the lower end of the crosstable. Similarly, this often applies to those towards the top who want to maintain their position and are therefore often inclined to split the point among others in a similar position.

In smaller tourneys (Altibox for example) we see more 'fight' - more decisive results. Why? It is the tournament structure. In this case a win in STANDARD PLAY nets you 3 points. If you have to play the ARMAGEDDON (10 min vs 7 min), a win only nets you 1.5 pts and a loss 1 pt. This is a serious attempt to get decisive results and it seems to work. Perhaps a mathematician can quantify this, but the variability in total points with this format between small/short tournament vs and larger/longer tournaments might need to be adjusted/tweaked.

Whatever the case, here we see an attempt to weaponize the point per outcome as a way to induce decisive results. However, 2 hrs + 10 sec is arguably TOO SLOW for this day and age...especially when so few games are being played in each round. Perhaps G/90 (or 75) + 5? It certainly is too slow for small tournaments. Here only 3 games were going at one time...commentators really did not have much to do and spectators in chat seemed to lose interest. So, Altibox as a spectating event clearly faltered on that count.

My personal favorite "solution" is to bring the self-interest of every single player into every game that is played and do so in a very tangible way: AWARD THE BULK OF $$ AT THE END OF THE TOURNAMENT BASED ON WINS. With this approach, you may have - lets say a $100,000 tournament with $40,000 divided up traditionally by your ranking at the end....but the balance of the prize fund would be awarded BASED ON THE NUMBER OF WINS you have at the end of the tournament. IF you chose to use these armageddon determiners, you might well want to weight a 'win' there for lesser $$ to further incentivize wins in standard play. Again, a mathematician might find a specific sweet spot in pay-out structure, perhaps due to the overall number of games played, but the idea itself seems valid and in accord with basic human nature to maximize your own chances for a better payday.

With such a structure (and less "generous" time controls), even the tail-enders have something tangible to fight for right up to the very end of a tournament. Furthermore, if the top end chooses to 'coast' late in a tournament ...they do so with the knowledge that they are letting others who score decisive results cut into their ultimate final pay-out.

I personally find this a fascinating idea at reducing draws in top level play, while keeping the traditional game alive and increasing the satisfaction for viewers. Thoughts?

Posts: 2004
Joined: Sat Jun 02, 2012 12:13 am

Re: Draw problem? Incentivize wins!

Post by carldaman » Sat Oct 17, 2020 10:25 pm

It's probably worth a few experiments. It can be introduced without affecting the current rating system. You would just reward wins monetarily. I'd like to see some organizer give it a go. Of late, there have been some tournaments that award 3 points for a win and 1 for a draw, so another step in that direction, in that people who win more games can finish higher as well.

Post Reply