Idea: Add a statistic to a player's profile showing an rolling average rate of moves per game per day.
Rationale: How frequently a player is likely to move is a real consideration for selecting/accepting an opponent. Time controls are an upper limit, but within time limits there can still be quite a range of move frequency player-to-player. Currently, there is a self reported "move frequency" setting, but this may not be accurate, may not be updated by the player over time, etc., etc.
Specific idea: Within the players stats box, calculate a statiistic that captures the true move frequency over some recent period. for example, this might be calculated as: average moves made per day over last 30 days/average number of games active in last 30 days. This will give at least give an approximate idea of how many moves a player makes per game per day.
Cons: Granted, in reality some games may be favored by a player while other sit and wait, and an average-based statistic cannot reflect this variability. Also, for players carrying only a few games, the statistic may be as much driven by their opponents move frequency (i.e., a non-sub with 3 out of 6 slow moving opponents will have a lower move frequency statistic). Nonetheless, I think that on average, a statistic like this would give a prospective opponent at least a general idea of what kind of move frequency to expect (better than the self-reported move frequency). Finally, the statistic I am proposing can currently be calculated/estimated from the current statistics as: number of moves this month / number of days so far this month / number of active games. So, the info is in some sense already there - but it is inconvenient to do the math for this every time.
rec'd with minor modifications:
Three (3) stats:
Average rate of moves/day for the last 30 days
Average for the last 180 days
Average for entire membership
This would show recent rates and historical trends, maybe
How about even a move rate chart, much like the green rating chart? how cool would that be!
Originally posted by gedwinI think this is a great idea. The current move frequency setting is basically useless and usually a misrepresentation of actual move frequency.
Idea: Add a statistic to a player's profile showing an rolling average rate of moves per game per day.
Rationale: How frequently a player is likely to move is a real consideration for selecting/accepting an opponent. Time controls are an upper limit, but within time limits there can still be quite a range of move frequency player-to-player. Currently, th ...[text shortened]... o is in some sense already there - but it is inconvenient to do the math for this every time.
Originally posted by Very RustyTruly, I am very happy with what we have. That is why I have been a subscriber for 3+ years and plan to continue to be.
Is there no end?
Will we ever be happy with what we have?
If suggestions of possible additions to site somehow offends you, I'd suggest that you avoid perusing the "Site Ideas" forum! 😛
I would like to have a Game Load Index in the statistics. Like an average number of moves a member must do in a day in order not to risque any time out. This GLI compared with how many moves they actually do is a sign of how likely they are to actually have problems with their amount of games.
My GLI is, at the moment 16.7 moves a day if my opponents moves immediately, 8.3 moves a day if they are as slow as they are allowed. Time bank's not counted.
(Yes - more statistics!)
Originally posted by Very Rusty"Site Ideas" is the name of this forum not Site Demands, so thinking up improvements saves russ money. Russ won't have to contract a consulting firm to find ways to make rhp more competitive. Direct input from the consumer is what reputable marketing firm does.
Is there no end?
Will we ever be happy with what we have?
Originally posted by gedwinThis conversation has come up before, and it's still a useless number meaning nothing.
Idea: Add a statistic to a player's profile showing an rolling average rate of moves per game per day.
Rationale: How frequently a player is likely to move is a real consideration for selecting/accepting an opponent. Time controls are an upper limit, but within time limits there can still be quite a range of move frequency player-to-player. Currently, th ...[text shortened]... o is in some sense already there - but it is inconvenient to do the math for this every time.
The statistic number given will never match how often that player will move in a game you started.
Example: You have 10 games in progress. Perhaps you move every day in 4 or 6 of them. Perhaps there is one game you move every chance you get in. Perhaps you have one game you have not moved in 7 days, or 21 days.
There are other players who have 200 games. Perhaps they have one game they move every chance they get. Perhaps they have 30 games they only move every 14 days in.
There are endless examples of why this statistic is useless, it's pretty much a number meaning nothing at all.
P-
Originally posted by PhlabibitWell don't end there then. I'm still waiting for a good one. C'mon!
There are endless examples of why this statistic is useless, it's pretty much a number meaning nothing at all.
Seriously though (and I'm being serious) if we're going to have a "move frequency" setting like
we already do that people use (some of us, at least) to help decide whether or not to accept
open invites, etc. based on mutual move frequency, then why not make it more accurate than an
arbitrary user choice. Is there anyway to predict the exact time a user will move in a particular
game? No. In the grand scheme of the universe is this stat meaningful? No. But neither are the
new stats on the profile pages (white/black win %, etc.), and I think those stats are a great
addition to the site. Other chess sites have this stat and I like it. So there.
In fact we do have a few numbers already:
* number of games in progress
* number of moves current month (could be useless at the bgeinning of a month)
* number of games played (total number)
* number of games played during the last year
* number of games played during the last 90 days
!Note that this is rated games only!
So howI do the math:
number moves per month/day of the month: rough(!) estimation of moves/day
Moves/day by active games: rough(!) estimate of moves per active games.
For me currently:
moves this month: 2612 date: 20 means roughly 130 moves per day. games in progress: 90, so roughly 1.3 moves per game and day. Note that this is a rough estimate.
other estimting methods:
Assumption moves per finished games: 30
number of finished games last 90 days times 0.33 moves per game and day.
Example Ponderable: 317 games finished by 0.33 rough estimate: 104.6 moves/day
So the rough estimates are in sufficient accordance to estimate that my move frequency should be roundabout more than once a day.
If you want a more detailed picture you could look up my clans and find out that the Crazy Aquarists play 7day TO games preferably, so in a lot of games, the opponent will not move daily, so I move more in the others.
I think we have a lot of data already, and don't really need another statistical value.
Originally posted by Traveling AgainHow about this...
Well don't end there then. I'm still waiting for a good one. C'mon!
Seriously though (and I'm being serious) if we're going to have a "move frequency" setting like
we already do that people use (some of us, at least) to help decide whether or not to accept
open invites, etc. based on mutual move frequency, then why not make it more accurate ...[text shortened]... great
addition to the site. Other chess sites have this stat and I like it. So there.
You give me your move value (mine is zero, but soon will change. My OTB friend QUIT)
Anyway, my feeling is that whatever the number is you come up with isn't really a true reflection of what a user can expect when playing you.
Perhaps I'm wrong? Give the formula and let's see what you come up with for any 3 or 5 given active users here at RHP. How is this number useful?
My other considerations were time controls used, vacation taken, opening moves, long endgames... yada yada.
Originally posted by gedwinHave to agree with Phlabby this is a meaningless stat.
Idea: Add a statistic to a player's profile showing an rolling average rate of moves per game per day.
Rationale: How frequently a player is likely to move is a real consideration for selecting/accepting an opponent. Time controls are an upper limit, but within time limits there can still be quite a range of move frequency player-to-player. Currently, th ...[text shortened]... o is in some sense already there - but it is inconvenient to do the math for this every time.
Say a player agrees to meet up with a friend and play out two games each day, averaging 36 moves per game.
In addition they play 8 other games, moving once per day in these games.
Their rolling average moves per day is 8, but they will only move once per day on your game.
Another player has 100 games in progess and makes 4 moves in each game each day.
Their rolling average moves per day is 4 and you will get 4 moves per day from them.
Which opponent would you rather play ?