Chocolate_Rhythm
Debutant
11 and counting!
-
LIVE: Geelong v Brisbane Lions - 7:30PM Sat
Squiggle tips Cats at 54% chance -- What's your tip? -- Teams on Thurs »
-
AFLW 2024 - Round 4 - Chat, game threads, injury lists, team lineups and more.
Can Dallas go all the way
- Thread starter BEEG
- Start date
- Tagged users None
- Banned
- #102
I can, if you like, give you an answer to the question of "How likely is it that Dallas' TRUE win% in these games is <60% given the results?".
I'll do the sum tomorrow unless someone wants to step in and do it first.
- Banned
- #104
Basically it goes like this;
I say Dallas has a less than 60% chance of winning a game decided by 3 points or less.
You say Horseshit! and show me the stats.
Then I run some maths and it gives me, in % terms, how likely I am to be wrong.
Does that make sense?
Also, have you got the W-L for Dallas over that same period but for 3 points instead of 5?
- Banned
- #106
Did it for Dynasty Bulls, Dynasty Spurs and Dynasty Lakers. Under 60% for the first 2, 65%+ for the third. I.e. Stern Card is worth a cool 5-10%
Hahah yeah, I guessed you weren't talking about Kobe when you said there is a reason for the Lakers batting at >60%.
How do you work out 'True win%' as opposed to just using W-L?
How do you work out 'True win%' as opposed to just using W-L?
- Banned
- #108
You can't really work it out, but you can give a best guess, plus a range.
For example, if the Celts have gone 57-72 in one-possession games over the last decade, I can tell you with 99% certainty that their "True" win% is somewhere between 33%-55%.
Considering just how much a team changes over a decade, and just how big a range I've just given you (22% across!), it makes it almost impossible to measure team clutchness in any true sense. There's just not enough data in almost all cases.
- Sep 21, 2004
- 46,673
- 53,300
- AFL Club
- GWS
Ok just added em up.
Dallas record in < 3 point games since Dirk became our #1
That record extrapolated over 82 games in first bracket, actual record in second bracket
09-10: 7-2 (64-18) (55-27)
08-9: 10-4 (58-24) (50-32)
07-8: 6-5 (44-38) (51-31)
06-7: 11-3 (64-18) (67-15)
05-6: 11-5 (56-26) (61-21)
04-5: 7-4 (52-32) (58-24)
- Banned
- #112
Cheers.
Basically, you can say with 99% certainty that Dallas should win at least 56% of one-possession games.
If you want to be a bit more lax, you can say with 95% certainty that Dallas should win at least 59% of one-possession games.
That's the best we can do.
However, I can't imagine too many teams would have better numbers than that, so it's impressive.
Basically, you can say with 99% certainty that Dallas should win at least 56% of one-possession games.
If you want to be a bit more lax, you can say with 95% certainty that Dallas should win at least 59% of one-possession games.
That's the best we can do.
However, I can't imagine too many teams would have better numbers than that, so it's impressive.
- Banned
- #115
It basically means their "True" win% range is anywhere from 56%-84%. Variance is a bitch.
In any case, you'll like this. For one-possession games;
Dirk post-Nash: 52-23
Kobe post-Shaq: 40-40
At the sports stats conference going on right now, someone presented a paper about the perils of having the go-to guy always take the last shot. The above numbers stem very much from that same idea. It's a lot easier to defend only Kobe than it is to defend Dirk+Terry+random cutter.
- Sep 21, 2004
- 46,673
- 53,300
- AFL Club
- GWS
Assuming Dalas keeps playing the way they are the records will be very close. IMO their 50/50 games are all harder. we win ours, they lose theirs and records are basically even
Dallas remaining schedule
Probable Wins: 14
50/50: 3 (at home to Boston, Denver and Orlando)
Probable losses: 1 (play Portland twice on road; likely to lose one
LAL
probable wins: 13
50/50: 4 (@ Phoenix, @ OKC, @ Atlanta, @ SAS)
probable losses: 2 (@ Orlando, @ Denver)
- Banned
- #119
I'm a fan of Dirk setting a high pick for Terry, who then curls right towards the right wing, giving the inbounder two genuine quick-trigger options.
The funny thing is, Terry's catch-and-shoot is a million times better running right than left, yet nearly all defenses choose to go over the Dirk screen, which leaves Terry's right side nice and open.
It might be the best 5-seconds-or-less-left play in the NBA.
- Sep 21, 2004
- 46,673
- 53,300
- AFL Club
- GWS
The Mavs certainly can go all the way. Its looking like a dogfight between the Nuggets, Mavs and Jazz for spot 2. Despite the Lakers recent slump I can't see them losing the #1 spot, and still think they are big favourites over any challengers in a best of 7 series
- Dec 8, 2004
- 22,762
- 156
- AFL Club
- Melbourne
- Other Teams
- Post Count: heaps
- Banned
- #124
There's only like 2 or 3 games between LA and Dallas now though right?
Its almost certainly going to be a LA/Dal/Den/Uta top four, and I really can't see any of those teams losing a first round matchup (seeing as GSW won't be there to destroy Dallas anyway).
Really hope for the Thunder's sake they end up matching with Utah or Dallas, and definitely not Denver.
Yeah, the lakers are on 18 losses, Nuggets and Mavs on 21. I think the Mavs might close the gap, but the lakers will probably lose 3-6 games for the rest of the season, so the Mavs will have to go close to winning out to take top spot. They might win more games than the Lakers down the stretch, but I dont know that they'll close the gap. Not sure who has the tiebreaker, could be important
Similar threads
- Locked
- Poll
- Locked
- Poll
