PDA

View Full Version : Some statistical analysis for the "trade down" crowd



Pauly
02-05-2008, 05:45 AM
** Updated figures at top of page 2 **

I've done some number crunching on 1st round draft picks and their likelihood of becoming HOFers and Pro Bowlers.

I've chosen the period 1967 to 1987 (21 seasons). The start date is the start of the common draft and the end date allows for the majority of the players drafted in that period who will will enter the HOF to be inducted. For the purposes of the exercise I've included Bruce Smith and Rod Woodson as HOFers even though they haven't been officially inducted yet.

I've measured careers by the number of seasons played because I did not have good access to the number of games played for players from the '60s. This tends to inflate the average careers for QBs in particular as a number of them had long careers spent mainly as backups.

I've arbitrarily called anyone who has a career of 3 seasons or less a bust, although a better mark would have been a benchmark of games started, but since I didn't easy access to those numbers the seasons proxy will have to do.

The Average PBs is the average number of PBs played per probowler. For example players drafted #1 overall from 1967 to 1987 played in 67 probowls, with 15 different players elected to the PB, which averages out at 4.5 probowls per probowler.

As I only had good access to information on 1st round picks I have only factored in rounds 1-26 of the draft.

#1 Overall:
HOFers: 7
PBers: 15
Average number of PBs played: 4.5
Average Career: 10.9 Seasons.
Busts: 0

#2 Overall
HOFers: 4
PBers: 11
Avge PBs: 4.9
Avge Career: 10.3
Busts: 0

#3 Overall
HOFers: 1
PBers: 12
Avge PBs: 3.6
Avge Career: 10.0
Busts: 0

#4 Overall
HOFers: 5
PBers: 12
Avge PBs: 5.2
Avge Career: 10.2
Busts: 0

#5 Overall
HOFers: 1
PBers: 11
Avge PBs: 2.8
Avge Career: 9.3
Busts: 2

#1-5 (As a Group)
HOFers: 18
PBers: 64
Avge PBs: 4.2
Avge Career: 10.1
Busts: 2

#6-10
HOFers: 8
PBers: 47
Avge PBs: 3.6
Avge Career: 8.3
Busts: 14

#11-#15
HOFers: 3
PBers: 29
Avge PBs 3.5
Avge Career: 7.3
Busts: 27

#16-#20
HOFers: 5
PBers: 31
Avge PBs: 3.3
Avge Career: 7.5
Busts: 22

#21-26*
HOFers: 2.5
PBers: 26.7
Avge PBs: 2.5
Avge Career: 7.3
Busts: 17.5

*(rounds 21-26 numbers have been reduced to account for the their are 6 picks in that bracket)

What the figures show is that you 14 times more likely to draft a HOF quality player with the #1 pick overall than you are if you are picking in the 20s.

FINOMINAL
02-05-2008, 09:34 AM
I can’t dispute your numbers, if they are accurate or not, but there were no busts at the number one, two, three or four pick in those years? I guess Ryan Leaf was 1998? Mandarich was 1989? Nice work though, I guess we will be getting a ten year starter in this draft, which will be great for us.

MrTree
02-05-2008, 09:40 AM
I can’t dispute your numbers, if they are accurate or not, but there were no busts at the number one, two, three or four pick in those years? I guess Ryan Leaf was 1998? Mandarich was 1989? Nice work though, I guess we will be getting a ten year starter in this draft, which will be great for us.


It's not to say there aren't busts...there definitely are. But from a pure statistical standpoint there are many less busts at #1 and many more extreme success stories. The only thing is that when a player has a HOF career from #1 (John Elway) you think "well yeah that's why he was drafted so high." But when a player busts (Leaf, Kijana Carter) it sticks out like a sore thumb because they went so high.

It is kind of like holding pocket Aces preflop in Texas Hold 'Em. You are at least 85% to beat ANYTHING. But people always remember the hands they lose with them (because you aren't supposed to lose with them) and also you never lose small with them. Holding the #1 pick is like holding pocket aces. The odds are heavily in your favor...but you better hope things don't go wrong or it will cost you.

Vendigo
02-05-2008, 10:39 AM
Nice work. One could, however, point out that 6 out of 21 players drafted #1 overall didn't even have a Pro Bowl appearance, meaning that 28% of all players drafted at #1 ended up being busts. That's a rather scary number if you take into account the salary cap issues of drafting there. The odds are still heavily into your favor, but do you bet the house on a 72% chance?

MrTree
02-05-2008, 10:47 AM
Nice work. One could, however, point out that 6 out of 21 players drafted #1 overall didn't even have a Pro Bowl appearance, meaning that 28% of all players drafted at #1 ended up being busts. That's a rather scary number if you take into account the salary cap issues of drafting there. The odds are still heavily into your favor, but do you bet the house on a 72% chance?

Using the poker analogy from earlier if you know you are 72% then you push all in given the opportunity preflop. Those are great odds.

Then you pray it works out.

I think the more important point from the original post is trading down is not so simple as others here make it out to be. We HAVE to get value to move off of the #1 pick. It seems like a lot of people around here are willing to throw away the value just to get an extra pick. That would be a disastrous decision for this team.

emocomputerjock
02-05-2008, 10:50 AM
Nice work. One could, however, point out that 6 out of 21 players drafted #1 overall didn't even have a Pro Bowl appearance, meaning that 28% of all players drafted at #1 ended up being busts. That's a rather scary number if you take into account the salary cap issues of drafting there. The odds are still heavily into your favor, but do you bet the house on a 72% chance?

Heck yes you do.

WISfinfan13
02-05-2008, 10:58 AM
Why did you choose a 1967-1989???? Why not just choose the last 21 seasons to do your analysis?? Scouting is different now than it was in the 70's...And I think it gives a more accurate evaluation of the point your trying to make..

Vendigo
02-05-2008, 10:58 AM
Those are great odds.


Yes, they are. But are they really great enough to bet your house on? That's what it ultimately boils down to: To stick with your analogy, I'll certainly push all in there, but I'll certainly don't bet my bankroll. One could seriously pose the question whether or not that's a limit we can afford playing.



It seems like a lot of people around here are willing to throw away the value just to get an extra pick. That would be a disastrous decision for this team.


I agree. I really don't get why people would even consider trading away the #1 for Dallas' two firstrounders. I would at the very least want an additional 2nd plus a starter on defense, likely more.

MrTree
02-05-2008, 10:59 AM
Why did you choose a 1967-1989???? Why not just choose the last 21 seasons to do your analysis?? Scouting is different now than it was in the 70's...And I think it gives a more accurate evaluation of the point your trying to make..


I'm guessing he cut it off because the players after 1989 are still current and for the most part are not HOF eligible. There are some surefire HOFers in those years too such as Manning.

WISfinfan13
02-05-2008, 11:00 AM
Plus if your picking first overall I personally expect a probowl type player...Anything less in book should be considered a bust. So 6 guys picked first overall would be busts in my book.

MrTree
02-05-2008, 11:04 AM
Yes, they are. But are they really great enough to bet your house on? That's what it ultimately boils down to: To stick with your analogy, I'll certainly push all in there, but I'll certainly don't bet my bankroll. One could seriously pose the question whether or not that's a limit we can afford playing.





I agree. I really don't get why people would even consider trading away the #1 for Dallas' two firstrounders. I would at the very least want an additional 2nd plus a starter on defense, likely more.

The odds are dictated by the game. In this case you just don't get better odds. It's a good calculated risk. To move off of it you have to feel you get a deal that represents a better calculated risk. In other words since the chances of getting a pro bowler/HOFer out of a later pick decrease by around 400% you have to see a similar decrease in what you lose if you bust the pick. (Also factoring in you get multiple chances to hit as well.)

I'm not opposed to trading down...but it HAS to work for us.

I think a lot of people fail to realize the tremendous impact hitting a homerun with the #1 would have for this team. It could be the kind of effect that pushes us back to contention in the big game.

2413fanphins
02-05-2008, 11:19 AM
Plus if your picking first overall I personally expect a probowl type player...Anything less in book should be considered a bust. So 6 guys picked first overall would be busts in my book.


I agree to a point, but how many qb's (just picked a position), go to the probowl?

one starter and three reserves... so are you saying you wouldn't be happy if you drafted a qb #1 overall, and he was the fift best qb in the afc?

it's not exactly going to be easy to put up better numbers than peyton manning, tom brady, and carson palmer.

please tell me there is no way you would consider the fifth best qb in the afc a bust?????

just asking

MrTree
02-05-2008, 11:23 AM
Plus if your picking first overall I personally expect a probowl type player...Anything less in book should be considered a bust. So 6 guys picked first overall would be busts in my book.


I don't evaluate on pro bowls. I evaluate on guys who help a team win games.

Vendigo
02-05-2008, 02:35 PM
In this case you just don't get better odds


That doesn't make them great, however. It just happens to be the only game in town. I'm not saying that you shouldn't draft at #1 - heck, I'm all for it. But pointing out that you have a one in three chance to mess up your salary cap for good (and hence mess up the possibility to turn this thing around for the next couple of years) is a legitimate point, in my view. And mind you, we didn't even talk about how many of those Pro Bowl players were actually worthy of a #1 selection. I assume that you are very aware of the fact that a Pro Bowl appearence doesn't make a great player. Vince Young made a Pro Bowl. I'd much rather shoot myself in the foot than shell out that kind of money for him.



I think a lot of people fail to realize the tremendous impact hitting a homerun with the #1 would have for this team. It could be the kind of effect that pushes us back to contention in the big game.


I agree. That's one of the reasons I'm all for staying put at #1 (provided, of course, that no one makes an offer we can't resist). Getting one truly great player beats getting two good ones. I merely think that you don't have that great of a chance of finding one that those numbers suggest at first. I'd still go for it, in the end.

Pauly
02-05-2008, 03:36 PM
Just some notes on the data. The point of the exerciese for me was to compare HOFers to busts.

Extending the data beyond '87 brings in too many players who are eligible for the HOF who haven't made it yet. I've deemed 2 players HOF (Bruce Smith, Rod Woodson) even though they haven't made it yet and there are another 4 or 5 players in the sample who stand decent chances of eventually getting into the Hall.

I know that counting probowls isn't an infallible measure of quality, but it is an objective criteria that can be validly compared across eras. The same for my definition of busts. Without games started/played data I had to use something that was objective and would be as valid for #1 as it is for #26.

The interesting facts for me:
- That the #4 overall is much better than the #3 overall on average.
- There were a surprising number of players who had more all pro selections than probowls and a few players named to "all decade" teams without playing in one probowl.
- The difference between picking in the #11 to #26 range is surprisingly small.
- The top 10 picks are genuinely much better than later picks (I always thought of this as just a lazy media myth).

Pauly
02-05-2008, 03:42 PM
Why did you choose a 1967-1989???? Why not just choose the last 21 seasons to do your analysis?? Scouting is different now than it was in the 70's...And I think it gives a more accurate evaluation of the point your trying to make..

I don't think scouting for 1st round picks has changed that much over the years - it's almost entirely high profile players from high profile colleges. It's the later rounds that the scouting has improved on.

Roman529
02-05-2008, 04:28 PM
Good work. I just think from the stats I have seen that choosing a QB with a first rounder is very risky.....you probably have a 50/50 shot of them being a franchise QB versus them being a bust. That is why I would rather go with a guy like Derek Anderson who has already shown he has the arm/skills to be up there with the top QB's.

Pauly
02-06-2008, 10:52 PM
**UPDATE**

People have wanted to see if the data changes with inclusion of later periods. I have extended the data to include the period 1967 to 1996 (30 years).

I have deemed any player with 5 or more probowl selections to be a HOFer, on the theory that those players will eventually be inducted. There are about 5 or 6 players from the original data who I have upgraded to HOF status by using this deeming.

I will not extend the data any further forward as I was aiming to review completed careers, and I have had to make some adjustments to career lengths to account for players still playing.

#1 Overall:
HOFers: 9
PBers: 19
Average number of PBs played: 4.3
Average Career: 11.1 Seasons.
Busts: 0

#2 Overall
HOFers: 9
PBers: 16
Avge PBs: 4.7
Avge Career: 10.3
Busts: 0

#3 Overall
HOFers: 4
PBers: 19
Avge PBs: 3.7
Avge Career: 10.2
Busts: 0

#4 Overall
HOFers: 11
PBers: 19
Avge PBs: 5.3
Avge Career: 10.2
Busts: 0

#5 Overall
HOFers: 3
PBers: 15
Avge PBs: 3.5
Avge Career: 9.8
Busts: 3

#1-5 (As a Group)
HOFers: 36
PBers: 88
Avge PBs: 4.3
Avge Career: 10.3
Busts: 3

#6-10
HOFers: 17
PBers: 65
Avge PBs: 3.8
Avge Career: 8.8
Busts: 15

#11-#15
HOFers: 14
PBers: 49
Avge PBs 3.3
Avge Career: 7.7
Busts: 27

#16-#20
HOFers: 10
PBers: 43
Avge PBs: 3.6
Avge Career: 7.8
Busts: 27

#21-26*
HOFers: 6.67
PBers: 40
Avge PBs: 2.6
Avge Career: 7.5
Busts: 23.3

*(rounds 21-26 numbers have been reduced to account for the their are 6 picks in that bracket)