| |
BGonline.org Forums
Making an error to lower your PR--How Stick can prove his point
Posted By: Phil Simborg In Response To: Making an error to lower your PR (Stick)
Date: Wednesday, 29 March 2017, at 12:54 p.m.
You would have made a big mistake playing on here in our chouette because you might well get a take or two depending on who was in the game and where they were on the sheet.
On a more serious note, I am not entirely convinced that the way the bots measure PR is necessarily the "best" way to measure a person's performance.
For example, as Kit and others have pointed out to me, you can make a relatively small mental error on a cube decision and get dinged in a huge way on your PR, but a similar mental error on checker play won't cost you much. PR measures what happens to match equity, it doesn't truly measure the players ability to think properly about how to play the game.
Also, positions like the one you show here create problems. What if you decide to play on and you are wrong. You have made a single mental error of judgment, but every roll you get dinged on your cube decision. One mental error, but 4 or 5 reductions in PR.
Also, the bots do not differentiate between errors at the beginning of the match and the end of the match. You get dinged the same amount for a wrong opening move on game 1 as you do in game 12. But in game 1 you might be much more correct to take some chances to "test" your opponents responses and strategies, because if you are wrong, you have plenty of time to adjust your game later, and so many more things (rolls) will happen before the end of the match, a .020 error will mean very little. But that same .020 error on the last game is more likely to cost you the match and might be a very bad gamble.
And lastly, the bots PR's are based on the concept that the bot is always right. We know it isn't, especially without a full rollout.
Now, I am NOT saying the PR isn't a decent guide to telling us about player skill....it's the best tool we have right now. But I am agreeing with Stick, and with Falafel, and others who say that we are putting too much emphasis on PR to the point where it can actually take away from the player's ability to win.
I remember years ago, when Stick lived with me, and Snowie first came out, that Stick said to play like Gammon Go and Gammon Save almost any time you are ahead or behind in the match, not just when it was pure gammon go and gammon save. Snowie did not agree with Stick, and he said that Snowie was full of shit.
Today, XG agrees with Stick, and even at 5A/4A and 4A/3A XG plays pretty much like GG and GS. How do we know what the next generation of bots will say, not just about opening moves, but about PVP plays and playing on for the gammon plays etc. etc.?
Now the flip side: too many players use the above and other arguments to explain away bad plays and bad PR. And I don't mean Stick. Stick pretty much knows when he makes a non-PR play. But I have heard many players at lower levels than Stick try to justify a really bad play or bad PR by saying they were "playing the opponent" and that's why they made a dumb cube or stupid play.
Now, to Stick personally. I think you've made your point, but you haven't really proven it. If you want to prove your point, that your PR's are not as good as they could be because you are playing to win, you can do that easily. Play 10 matches publicly with the sole purpose of playing at the lowest PR you can, and let's see what your PR is. If you can play under 3 PR, or even as low as 2.5, then, from then on, when your recorded matches come in 1 or 2 PR higher it would be very easy to believe that your higher PR's are because you are intentionally adjusting for AtO.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.