On a different note, since I have the lowest H[ugh]-rating out this discussion, by the Hugh Hypothesis that means I play better opponents as my global ranking/rating is high. I am not sure that I agree since I mostly play players who have participated in this discussion (which in turn I am assuming means you guys and myself are playing with the same group of players). Thus, I must just be a lessor player and the Hugh Hypothesis cannot be true.
That is, I am not sure that a high Global Ranking and low h-rating implies anything about the quality of your opponents. I think it really just corresponds to you having a good week/month making the volatile global ranking jump when really your h-rating is showing how good you are.
That is h-rating is global, global ranking is local in the sense of time.
Mongrel wrote: Suppose I have 3 balls, one red, one blue, one green
Yertle wrote: perhaps just hit backspace instead of using strike through that can still be read (/Y edit>
Well that totally defeats the purpose of introspect and sharing.
If I really cared that anyone saw me use potty language, I'd tell them to turn on the profanity filter and leave me alone. Then, I've done them the favor of protecting them from all future profanity by exposing them to a little.
Mongrel wrote: Suppose I have 3 balls, one red, one blue, one green
I love it
(8-Y)
Re-inserting some language from my first write-up into M-57's last one. That i think better explains the statistic.
I left the formula out of my original write-up since I know math confuses people, but the one below (if correct i lost track of that thread within this thread) is easily glossed over and could be left in.
---
The H-score is a statistic that expresses your win percentage with respect to the expected norm. Obviously, the larger the number of opponents you face in a game, the lower your probability of winning. The H-score takes this into account by normalizing or “Gamesize Adjusting†those larger games by taking the results of all of your games and applying the formula:
(#Opps*#Wins)/(#Opps*#Wins+Losses)
Because the stat is gamesize adjusted, a 50% H-score means you are winning 1/2 of your 2-player games, 1/3 of your three player games, 1/4 of your 4-player games and so on.
Even more concise. I like.
A score, huh? Nice try..
I loathe acronyms. This statistic should be H something as per asm's post #89.
Amidon37 wrote:I loathe acronyms. This statistic should be H something as per asm's post #89.
Agreed on both counts.
I'm happy with the description as it has evolved in the thread. I'm not sure I like the gloss over with the formula. With those who want to know how it is calculated, too many details are left out, and it does nothing for those who just want to know what it does. Here is an edit (feel free to edit!):
---------
The H-score is a statistic that expresses your win percentage with respect to the expected norm. Obviously, the larger the number of opponents you face in a game, the lower your probability of winning. The H-score takes this into account by normalizing or “Gamesize Adjusting†those larger games so that a 50% H-score means you are winning 1/2 of your 2-player games, 1/3 of your three player games, 1/4 of your 4-player games and so on.
As for score vs rating vs ranking: I just noticed that what I've been calling "Global Rating" is referred to as "Global Ranking" on profiles, but if you click on the Rankings tab it is referred to as "Score". Another option is Percentage (or % for short), which may seem weird as it is an adjusted percentage, but so is baseball's "slugging percentage" (which doesn't even range from 0 to 1!)
The H-score is a statistic that expresses your win percentage with respect to the expected norm. Obviously, the larger the number of opponents you face in a game, the lower your probability of winning. The H-score takes this into account by normalizing or “Gamesize Adjusting†those larger games so that a 50% H-score means you are winning 1/2 of your 2-player games, 1/3 of your 3-player games, 1/4 of your 4-player games and so on.
---
small edit for consistency in last sentence.
Ha! Took me a while: "three player" got changed to "3-player". Agree.
I think the description and example scores are presented in such a nice and compact manner that it would be nice to include the formula, which really isn't that difficult to understand, and also provided that it can be presented concisely and in a manner that makes it reasonably easy to understand.
=====
The H-score is a statistic that expresses your win percentage with respect to the expected norm. Obviously, the larger the number of opponents you face in a game, the lower your probability of winning. The H-score takes this into account by normalizing or “Gamesize Adjusting†those larger games so that a 50% H-score means you are winning 1/2 of your 2-player games, 1/3 of your 3-player games, 1/4 of your 4-player games and so on. It is calculated as follows:
{Sigma(W*O)}/{Sigma (L+(W*O))} where:
W = Number of wins in that category of #-player games.
L = Number of losses in that category of #-player games.
O = Number of opponents you would play if you played one game in that category.
=====
I tightened up the variable definitions just a tad. The least concise word in them is the word "that". Replacing it with "each" would be more confusing, but replacing it with "each individual", may be the best choice though it seems a bit wordy.
#152 is good on wording, i still like the correct formula (whatever that may end up being) is fine to have in there. but i'll defer to whatever you decide.
anyone else have an edit?
Actually, that is nicer than before because the formula comes last. I'm happy with that, but I'm a math person. Perhaps Sigma() is replaced by Sum()? Yertle, does that look too out of control for the Help file?
i think "that category" is ok, can't think of anything much better...
Number of wins in a specific category #-player games.
Number of wins in #-player-games.
Number of wins in all #-player-games.
Number of wins in each #-player-games.
or change the formulat to be:
{Sigma(W2*O2)+...+(W#*O#)}/{Sigma (L2+(W2*O2))+...+(L#+(W#*O#))} where:
W# = Number of wins in #-player games.
L# = Number of losses in #-player games.
O# = Number of opponents you would play if you played one game in that category.
Awesome! I have added it to http://www.wargear.net/help/display/Rankings which has post 154 and the example from 109.
Seems to make sense to get rid of G rating and just utilize the H rating right? That way the G rating information can taken out of the Help and I don't think it will be overly crazy on the page. It may be long with the formula and the example, but if people just want a basic understanding then they can stop at the description of the H rating (will update if the name changes, but left it as to what it is currently on the tables), but if they really want to know and try and figure it out then there is the formula and the example to edge them along.
Thanks!
weathertop wrote:
or change the formulat to be:
{Sigma(W2*O2)+...+(W#*O#)}/{Sigma (L2+(W2*O2))+...+(L#+(W#*O#))} where:
W# = Number of wins in #-player games.
L# = Number of losses in #-player games.
O# = Number of opponents you would play if you played one game in that category.
That's why I added the example to the Help page, from Sigma or Sum I don't necessarily get that it includes all categories of # of player games. That formula does help a bit I think, although, IMO, the example is still valuable and then we're just left with a longer formula too.
Great! Now it needs to be actually expressed as a % on the Rankings page and I think it would make a great deal of sense to move it directly adjacent and to the right of the Win% column. Kinda like putting that slugging percentage right next to the batting average.
..although it does occur to me there might be some good reasons to leave it next to the WG Score. I just don't think they're as strong.