Total wins

[Fluffy]

With the roosters looking like getting their 1000th win in 08 i thought id see where the other teams rank.

Roosters 999/1898
Souths 946/1831
St George 910/1545
Balmain 871/1705
Canterbury 790/1522
Wests 734/1691
Manly 731/1280
Norths 687/1665
Parra 619/1340
Newtown 583/1305
Cronulla 461/950
Penriff 387/935
Canberra 333/625
Brisbane 309/478
Newcastle 244/465
melbourne 140/240
Warriors 135/291
St George/Illawarra 112/213
Cowboys 101/291
Wests Tigers 72/179
Titans 1/3

Incedently the title for most losses goes to the bears with 916

 

[Crushercleal]

How good are we - we only came into the Comp in 1947, and we already have 3/4 of the Rooters wins, and they were there 40 yrs before us.

 

[Fluffy]

we are 4th behind brisbane, stgeorge and melbourne

 

[Dan]

you should exclude any team that has played less than 1000 games to make a satisfactory sample

 

[Fluffy]

it will not change the total wins.

To start excluding for % wins you are better off just grouping teams who started at the same time.

ie manly vs Parra

 

[Dan]

nah I am talking of a correct sample group

a team that has played only 200 games ahs not had the opportunity to win or lose as many times. As time goes on that margin will increase or decrease.

This is why a sample gorup has to be a minimum of 1000 games!

TO be even more correct you should really say then base it on the first 1000 games only.

But the basic point is if you take the same sample size for each team you WILL get different resutls

 

[fLIP]

Take percentages.

 

[Matabele]

The parra/manly comparison is priceless. Tools.

 

[Dan]

Take percentages.

still doesnt give you an accurate picture.

 

[fLIP]

Your never going to get an accurate comparison. Even by taking just the 1000 games played teams.

There are so many variables throughout a teams history.

 

[Dan]

but as it stands the stats are completely useless. At least with the same sample size you will get a more accurate result. statistics will never ever be 100% accurate but as they stand in this example its not even worth reading

 

[Fluffy]

very true Flip,

Teams who played in the early days had far different rules, ie local players only etc compared with today. The saints of the 50's and 60's didnt have the salary cap.

and of coarse as Dan says many teams havnt played enough games to get a good sample size. In pure mathamatics its generally a minimum of 2000 for something to be considered statistically true.

Hence why i feel comparing like with like is the closest you can get. All the foundation teams played 80% (newtown) or more of the time together under the same rules etc, manly and parra both started in 47, dogs and saints were a little earler but all of simlilar eras.

 

[Fluffy]

Dan they are worth reading for 3 reasons:

1 Manly's total wins for those who would like to know
2 Manly Vs Parra win ratio difference
3 Norths have the most losses

 

[fLIP]

The point dan is trying to make is that the longer a team playes the more stable their stats will be.

Take the titans. they currently sit on winning 33.3% of their games. But if they win on the weekend it will change to 50%.

what he's saying is its unfair to judge two teams together when ones stats are not as indicative as a team with more games under their belt.

 

[Dan]

but meaningless fluffy

 

[Fluffy]

you said worth reading

and meaningless could be said for many things posted on here.

Flip i agreed with Dans reasoning hence why i said only teams with similar years can be compared to each other.

 

[Matabele]

Teams that have a city of 3 million to themselves should be excluded.