Wins and lottery position: recent history, for perspective

[Boris Badenov]

Inspired by PhoSita's wins projection post, and also by the many discussions about exactly how bad we'd need to be to get a good pick, I though I'd look at where different numbers of wins would put you in the lottery.

I went back and looked at the lottery position 20, 25 and 30 wins would get you for every year since 2000.

The first number shows the average lottery position with that number of wins, and the parentheses show the high and low picks.

Wins   Average  (High, low)

20     2.5      (1st, 4th) 
25     4.8      (2nd, 7th)
30     7.8      (4th, 10th)


Note that these are *lottery* positions. Actual draft positions (which I may put together later) are going to be more spread out due to the randomness of the lottery (though the averages may not be all that different).

Also, the trend in recent years has been toward more bad teams, meaning it's gotten tougher to get a good lottery slot.

Bottom line: 30 wins is likely to get you a pick in the 5-10 range. 25 wins will get you a pick in the 3-6 range. And 20 wins almost guarantees you something in the top 3 range.

Here's Pho's OP:

(http://forums.celticsblog.com/index.php?topic=66228.msg1508536;topicseen#new)

[rondoallaturca]

Thanks for the post. I think it's worth pointing out that unlike other drafts, many teams will be tanking for 2014, so we're probably looking at the upper end of the averages.

Regardless, it's still nice to know that we can still secure a 5-10 pick even with Rondo on board, which serves as more reasoning why there's no incentive for Ainge to move Rondo.

[PhoSita]

TP, thanks for doing this.  Definitely accords with what I would have assumed.

As you pointed out, though, the "competition" for being bad has gotten stiffer in recent years, and the 2014 draft is giving a lot of teams incentive to be terrible, so it will be even harder.

The team we have now might outperform the projections in my Win Shares post if only because so many of their opponents would be even less talented.

If the Celtics really want to get a top 5 pick there's still (demolition) work to be done this summer.

One quibble -- the best "guarantee" you can get in the lottery is 1-4, so even 20 wins won't guarantee you top 3.  However, most years 20 wins or fewer will pretty much assure you of having the worst record. 

This year, though, I wouldn't be surprised if we see some teams challenging for the futility record (i.e. closer to 10 wins than 20).

[LooseCannon]

I'd rather not do any demolition work and just hope that the Celtics injury trend continues. 

[fairweatherfan]

Nice post.  If you're looking for more nuance, you could probably do a simple regression of games won and draft position, rather than looking at round numbers of games won.

[Boris Badenov]

Thanks for the post. I think it's worth pointing out that unlike other drafts, many teams will be tanking for 2014, so we're probably looking at the upper end of the averages.

You know, I expected the same thing, but I couldn't see any strong relationship between draft strength and how many bad teams there were. As examples, in the Oden/Durant draft the worst team had 22 wins, and 26 wins would have put you in 3rd. 2009 was a mediocre draft but 20 wins would have put you 4th and 25 would have put you 7th.

[Boris Badenov]

Nice post.  If you're looking for more nuance, you could probably do a simple regression of games won and draft position, rather than looking at round numbers of games won.

Yeah. The fit of the model would also tell you how random it is considering the influence of the lottery.

[droopdog7]

I thought this was going to be about the positions that won each lottery.  I know the worst team doesn't win that often but I am curious whether this information can be found anywhere?

But yes, this info is interesting too.

[tarheelsxxiii]

Nice post.  If you're looking for more nuance, you could probably do a simple regression of games won and draft position, rather than looking at round numbers of games won.

Yeah. The fit of the model would also tell you how random it is considering the influence of the lottery.

I appreciate this. I'm pretty old-school and anti-tanking, but this is kinda cool.

[CelticAZ]

I tend to believe that there will be a number of teams tanking and that 10-15 games won will probably end up putting a team in the 2nd half of the lottery.

This is going to be a ugly year for a number of teams/cities; conversely, it's going to be a glorious time for the true championship contenders.  Will we see another 70+ win season from some team? How many 60+ win teams will there be?

[Birdman]

Well we struck out on one draft pick already with the rebuilding stage..Kelly "Oh-No" Olynyk

[Boris Badenov]

Nice post.  If you're looking for more nuance, you could probably do a simple regression of games won and draft position, rather than looking at round numbers of games won.

Yeah. The fit of the model would also tell you how random it is considering the influence of the lottery.

I appreciate this. I'm pretty old-school and anti-tanking, but this is kinda cool.

You could go a step further and consider adding moderating or mediating variables to the model, too

I can't think of any right now, but this could be the start of a new career for you

One variable would be quality of the draft class - that would answer rondoallaturca's and PhoSita's question about whether there's more tanking when the field is strong.

I'm sure there's a better way to model it too, but I haven't given it a lot of serious thought.

[droopdog7]

Well we struck out on one draft pick already with the rebuilding stage..Kelly "Oh-No" Olynyk

Thanks Nostradamus.

[PhoSita]

Nice post.  If you're looking for more nuance, you could probably do a simple regression of games won and draft position, rather than looking at round numbers of games won.

Yeah. The fit of the model would also tell you how random it is considering the influence of the lottery.

I appreciate this. I'm pretty old-school and anti-tanking, but this is kinda cool.

You could go a step further and consider adding moderating or mediating variables to the model, too

I can't think of any right now, but this could be the start of a new career for you

One variable would be quality of the draft class - that would answer rondoallaturca's and PhoSita's question about whether there's more tanking when the field is strong.

I'm sure there's a better way to model it too, but I haven't given it a lot of serious thought.

One thing to keep in mind is that when a bunch of teams all try to get bad, somebody has to win the games.  So it's very difficult for there to be very many teams with fewer than 20 wins.

I think that you might actually tend to see more teams tanking in years subsequent to a really good draft.  E.g. when others teams saw how nicely tanking for the 2007 and 2008 drafts worked out for the Thunder, a lot of other teams began adopting that strategy.

[Celtics4ever]

The thing to really note is what kind of swings generation picks take after you win.   LeBron, Duncan and Bird all had huge turnarounds for their franchises.