[Poll] Points Inflation - Guess how much

[ppgangster]

Do you take into account the points of all of those inactive players?

If so I would go for over 200. If not, a lot less!

[a.sub]

If so I would go for over 200. If not, a lot less!

that reminds me of

Im gonna live forever or die trying

(catch 22 anyone?)

[lancehoch]

You guys are all forgetting the busted multis and point resets. Between *manimal, Ramned, and the like, I am sure that 25,000 points were removed from the system, and another 10,000 from point resets.

[john9blue]

I'm just using the same method pi did- looking at the scoreboard.

[Timminz]

I'm just using the same method pi did- looking at the scoreboard.

That's the only way that makes sense. If you counted every member to ever sign up, you'd get no inflation at all.

[e_i_pi]

I'm just using the same method pi did- looking at the scoreboard.

That's the only way that makes sense. If you counted every member to ever sign up, you'd get no inflation at all.

Exactly.

Also, keep in mind there are two ways of approaching this:
1) Sum up the entire scoreboard and divide by the number users
2) As above, but discount the new recruits scores

Why consider #2? Because only 7% of new recruits stay at the site, and 14% of the scoreboard is new recruits. So really, only 87% of the players on the scoreboard will still be there in a months time. The remainder will have played their 4 games, likely deadbeated most of them, then left the pool.

[e_i_pi]

You guys are all forgetting the busted multis and point resets. Between *manimal, Ramned, and the like, I am sure that 25,000 points were removed from the system, and another 10,000 from point resets.

It's a drop in the ocean compared to the new recruits that join, deadbeat, and never come back

[Timminz]

I'm just using the same method pi did- looking at the scoreboard.

That's the only way that makes sense. If you counted every member to ever sign up, you'd get no inflation at all.

Exactly.

Also, keep in mind there are two ways of approaching this:
1) Sum up the entire scoreboard and divide by the number users
2) As above, but discount the new recruits scores

Why consider #2? Because only 7% of new recruits stay at the site, and 14% of the scoreboard is new recruits. So really, only 87% of the players on the scoreboard will still be there in a months time. The remainder will have played their 4 games, likely deadbeated most of them, then left the pool.

How about only discounting 93% of new recruits?

[e_i_pi]

I'm just using the same method pi did- looking at the scoreboard.

That's the only way that makes sense. If you counted every member to ever sign up, you'd get no inflation at all.

Exactly.

Also, keep in mind there are two ways of approaching this:
1) Sum up the entire scoreboard and divide by the number users
2) As above, but discount the new recruits scores

Why consider #2? Because only 7% of new recruits stay at the site, and 14% of the scoreboard is new recruits. So really, only 87% of the players on the scoreboard will still be there in a months time. The remainder will have played their 4 games, likely deadbeated most of them, then left the pool.

How about only discounting 93% of new recruits?

I've done that already Timminz Doesn't make a great deal of difference to the overall, but in the interests of accuracy, it makes it closer to the mark

[a.sub]

well?!?!?!?!?!?!?!

[e_i_pi]

well?!?!?!?!?!?!?!

I'll reveal in about 48 hours time. I have the stats at work

[a.sub]

well?!?!?!?!?!?!?!

I'll reveal in about 48 hours time. I have the stats at work

your killing me PI

[Gypsys Kiss]

I worked it out to be about 14 points per person, but that was 2 weeks ago and I cant remember how I did it............

[Robinette]

Geee... this is simpler than i was trying to make it before...
As I look at it this morning, it looks quite simply to be about 64 points...

I can see no reason to discount for the new recruits who will fall off the scoreboard within 30 days as they will always be replaced by others. So this is essentially a 'snapshot' of the current points inflation...
well, my best guess anyway...

I am anxious to see the calculations you have come up with...
It should be very interesting... and likely even to cause a bit of a stir...

[jpcloet]

I took a 10 page sample and got 272......

[timmy1]

Do you take into account the points of all of those inactive players?

If so I would go for over 200. If not, a lot less!

Good point, although those inactives are the assumed reason for inflation. It would be interesting to see the periodic averages of all active players and how it has changed through time.

[e_i_pi]

RESULTS AT TIME OF COMPILATION

Total number of players: 22617
Total points on board: 26758896
Average points per player: 1183.13

Total number of new recruits: 3217
Approx number of new recruits that will leave: 2755.43
Approx number of players that will still be here in a month: 19861.57
Total points minus new recruits that will leave: 24069165
Average points per player, minus new recruits that will leave: 1211.85

So the poll was about right. It's 183.13 points per player, or 211.85 if you use the arguably more accurate method of discounting 6/7ths of the new recruits (as they won't be here in a month).

Interesting facts:

• About 8000 of the 22617 players are above average in points. This shows that there is massive skewing in the scoreboard due to points inflation. If inflation didn't occur, you would expect 11309 players to be of above average points.
• 1000 points, a median you would expect given no points inflation, is actually about 57.2% of the way down the scoreboard, not 50%
• Of the players of above average score, only one was a new recuit
• The Standard Deviation was 399.18
• 12% of the population were +1 or more StDev
• 23% of the population were 0 to +1 StDev
• 56% of the population were -1 to 0 StDev
• 9% of the population were -1 or more StDev

All that StDev talk might look like mumbo-jumbo, but the basics of it is this:

• Winner stay, losers leave (12% vs 9%)
• Almost 2/3rds of the population are below average
• When you start an account, you are in the 57th percentile
• Before leaving new recruit status, you are likely to drop to 975pts, which is the 66th percentile

Doesn't really promote people to stay on now, does it?

[jpcloet]

Only if points matter, I don't see "Fun" in the equation.

[e_i_pi]

Only if points matter, I don't see "Fun" in the equation.

And with the sweep of a single hand, the admin glibly swat away yet another plain piece of evidence suggesting that something ought to be done to administer the site. Bravo chap, bravo

[jpcloet]

Well the truth is points do matter to some. Does anyone have a suggestion?

[Timminz]

Only if points matter, I don't see "Fun" in the equation.

I definitely agree that points are not the be-all, and end-all, but if points didn't matter at all, there would not be any.

The way I'm interpreting this data, is that as time passes (inflation increases), new recruits are starting at an increasingly greater disadvantage.

In my opinion, something needs to be done to curb inflation. No Points For New Recruit Deadbeats, would be a large step towards this goal.

[Optimus Prime]

Only if points matter, I don't see "Fun" in the equation.

And with the sweep of a single hand, the admin glibly swat away yet another plain piece of evidence suggesting that something ought to be done to administer the site. Bravo chap, bravo

Oh wait... jpcloet is an admin... when did that happen? Oh right... he isn't. I remember now.

e_i_pi, despite your constant disapproval of how this website is run, it is administered on a daily basis. I spend an awful lot of time making sure that happens, it's my job, remember? Constantly bringing up you displeasure only serves to make yourself look bitter.

I'm one of the first to admit that there is point inflation going on, and at some point I'm sure us admins will get around to trying to tweak the system a bit, but the fact of the matter is that the point inflation isn't really the top priority right now as far as site improvements are concerned.

Keep figuring out ways to try and resolve the problem however, because the information will come in handy sometime, that I'm sure of.

Regards, OP

[e_i_pi]

Only if points matter, I don't see "Fun" in the equation.

And with the sweep of a single hand, the admin glibly swat away yet another plain piece of evidence suggesting that something ought to be done to administer the site. Bravo chap, bravo

Oh wait... jpcloet is an admin... when did that happen? Oh right... he isn't. I remember now.

e_i_pi, despite your constant disapproval of how this website is run, it is administered on a daily basis. I spend an awful lot of time making sure that happens, it's my job, remember? Constantly bringing up you displeasure only serves to make yourself look bitter.

Alright I tell you what, I'll stop being bitter when the admin start making site improvements... that's right, I said the admin. Apparently that isn't jpcloet, any of the tournament directors, mapmakers, moderators, or chatters. That's lack, yourself, and Andy - absence, warnings, and bananas don't count as site improvement FYI.

I'm one of the first to admit that there is point inflation going on, and at some point I'm sure us admins will get around to trying to tweak the system a bit, but the fact of the matter is that the point inflation isn't really the top priority right now as far as site improvements are concerned.

It's not a question of admitting that it's going on - it is going on. Admitting it's going on is like admitting the sun rose this morning.
If I remember correctly, the last 3 site improvements were:

• Strikethrough text, a 1 minute job
• Pink CC, a 10-20 minute job
• The return of normal CC, a 1 minute job

I'm just plain astonished that this counts as 'site improvement'. Oh wait, you got rid of two community managers as well. J'excuse.

Keep figuring out ways to try and resolve the problem however, because the information will come in handy sometime, that I'm sure of.

I figured out several ways, they're buried somewhere between page 5 and page 63 of the Sugs and Bugs forum.

Regards, OP

Regards, e_i_pi

[Optimus Prime]

If I remember correctly, the last 3 site improvements were:

• Strikethrough text, a 1 minute job
• Pink CC, a 10-20 minute job
• The return of normal CC, a 1 minute job

Don't forget about the rest of them over the past few months....

• New API to allow for discovery of new Tournament and Game statistics
• Tournament thread links inside gameplay area
• Round numbers included on My Games and Game Finder Pages
• Next Set Values being added in-game so as not to require BOB
• New Profile Fields

The problem is that you don't actually think that small things should count as site improvements. All you are interested in is the big bang type of stuff, which quite frankly isn't all that this place is about. Sometimes the small stuff is what needs to be done, not the big stuff.

[jbrettlip]

Remember that day it was pink??? That was awesome!!!