Here's a pretty understandable example for chess ratings - but obviously, GR are a bit different.

The statistical nature of the Elo system is tricky to handle, but a non-statistical mathematical analysis is available; it forecasts the same basic results in match and tournament play as the statistical method over differences up to about 700 Elo ratings points. Here's my method:



Take the difference in points between a player's own rating and the average of their opponents' ratings and call this P'. Then take 0.337046 (P x 1.063x10-4) and raise this to the power of (P/200). This will give you the lower rated player's score as a fraction of the higher rated player's score. It is then a very simple process to calculate the individual player's expected score; just remember to take the higher percentage share if the player is rated higher than their opposition, or the lower percentage if the opposition has the higher rating.

Is Ghost Rating even going to exist going forward?

Are we talking about a 5-way draw here, Indy?

Yeah, TGM will still the GR for us.
At one point, we should get someone to work with him so that he doesn't have to do it (unless he wants to).
Is there a way to program this right into the site?

"Is there a way to program this right into the site? "

That would be the best solution, perhaps with a few modifications. I would like to see ratings broken down by sub-type. It would be awesome if this site had a whole range of statistical options. I wish I was a programmer and knew how.

Yes, we are talking about a 5 player draw.

Computing GR changes for WTA games is fairly straightforward. Essentially, each player pays 1/17th of their GR points into a pot. The winner gets all of the GR points, while the losers get nothing. In the case of a draw, the players in the draw split the pot evenly.

For your above example, the pot of GR points adds up to
(300 + 150 + 110 + 120 + 101 + 120 + 103) / 17 = 59.0588235
Players A, B, C, D, E each get an equal share of that pot, while F and G get nothing.
Here are the resulting GRs
A 294.16
B 152.99
C 115.34
D 124.75
E 106.87
F 112.94
G 96.94

I thought it was 1/17.5 (2/35ths), not that it affects the outcome you did that much. And yebellz has it down well for WTA. The tricky part if computing it in PPSC when there is a winner (it's identical for a draw in both).

If that's actually then the case then, when you lose, the skill of your opponents doesn't matter - you lose the same amount. I didn't think this was the case, but I'm obviously mistaken.

@Yonni: well, in the above example, player A loses GR points despite being part of the draw. This is because his opponents' GRs are lower than his, and therefore 2/35ths of their GRs aren't as big as his.

So in this case, I contributed roughly 300 x 1/17 = 18 , and received 59/5 = 12 . You guys are wonderful, I GREATLY appreciate the info.

Not as concerned in a PPSC where there is a winner - and it wasn't me lol. THAT you can't stop. Agreeing to a draw...that's a choice.

Thanks again, Yeb, Draug, Yonni, TB. Greatly appreciate it.

@Yonni - The loss is always the same if you are defeated or survive a WTA game. Opponent skill only affects potential winnings in a win or draw (or PPSC survive) as not everyone contributes equally, unlike with points.

TB, good point about the draw. However, it's a shame that there isn't a mechanism more similar to ELO where you lose point in proportion to the level of your competition.