I think Alex's bid prices don't match up well with his computed projected prices.Before I continue, I have to clear up one misconception (or, more likely, omission) from Eugene's comment. Peter Kreutzer - a.k.a Rotoman - has been generating the projections for Patton & Company for the last few years. I can't say for certain how long he's been at the wheel. However, the fact that the projections are generated by Rotoman while the bids are generated by Patton is the greatest reason for the disparity. In reality, you're buying two different sets of opinions...even if you're not buying two separate projections.Several second basemen fit that profile - high average, high SB. Yet, the Patton bid isn't quite there. If he just bid his projections he'd probably get Brian Roberts. Same goes for (Howie) Kendrick b/c his projected BA is quite high with solid steals. (Robinson) Cano is another BA guy, but without the bags. (Placido) Polanco as well. I'd be interested in seeing the projected price versus AP Bid price.
Since Eugene commented on A.L. middle infielders, it's worth looking at them to see where the price differences fall.
Ten Most Expensive A.L. Second Basemen 2009
| # | Player | $ | Sal | +/- | AP | PK | 2008 |
| 1 | Ian Kinsler | $26 | $36 | -10 | $34 | $28 | $34 |
| 2 | Dustin Pedroia | $25 | $34 | -9 | $27 | $22 | $36 |
| 3 | Brian Roberts | $27 | $31 | -4 | $29 | $23 | $33 |
| 4 | Alexei Ramirez | $18 | $28 | -11 | $23 | $21 | $26 |
| 5 | Robinson Cano | $27 | $20 | +8 | $17 | $18 | $15 |
| 6 | Jose Lopez | $18 | $17 | +1 | $16 | $12 | $24 |
| 7 | Mark DeRosa | $10 | $16 | -7 | $15 | $16 | $22 |
| 8 | Placido Polanco | $17 | $16 | +1 | $16 | $15 | $19 |
| 9 | Howie Kendrick | $16 | $15 | +1 | $15 | $15 | $14 |
| 10 | Aaron Hill | $27 | $11 | +16 | $11 | $8 | $4 |
| Average | $21 | $22 | -2 | $20 | $18 | $23 |
A ha! I thought Patton's prices were conservative. Rotoman (represented by the PK column in the chart above) is even further away from the market than Patton is, missing the mark by $4 per player. His conservatism is most evident at the top of the chart, where last year's $30+ hitters reside. Rotoman is anticipating regression to the mean to the extreme. While he might be overly cautious, his caution on these hitters is warranted compared to the market on Kinsler and Pedroia. If you only looked at Rotoman's projections, you would have avoided losses on these hitters.
The down side is that you would had the same problem in an A.L. using Patton's prices and you might have missed out on all of these hitters. Rotoman doesn't get a single one of these hitters outright, and only ties on DeRosa and Kendrick.
Is Rotoman just stingy with middle infielders or across the board?
Ten Most Expensive A.L. Outfielders 2009
| # | Player | $ | Sal | +/- | AP | PK | 2008 |
| 1 | Grady Sizemore | $15 | $41 | -27 | $38 | $26 | $40 |
| 2 | Carl Crawford | $36 | $36 | 0 | $35 | $30 | $21 |
| 3 | Josh Hamilton | $11 | $35 | -24 | $30 | $27 | $38 |
| 4 | B.J. Upton | $19 | $35 | -15 | $34 | $25 | $31 |
| 5 | Nick Markakis | $22 | $33 | -11 | $30 | $25 | $28 |
| 6 | Jacoby Ellsbury | $36 | $31 | +6 | $29 | $31 | $33 |
| 7 | Ichiro Suzuki | $31 | $30 | +1 | $30 | $29 | $35 |
| 8 | Matt Holliday | $14 | $30 | -16 | $30 | $27 | $38 |
| 9 | Alex Rios | $16 | $29 | -13 | $29 | $25 | $33 |
| 10 | Jason Bay | $26 | $29 | -2 | $23 | $18 | $29 |
| Average | $23 | $33 | -10 | $31 | $26 | $33 |
Yikes. Rotoman is even more penurious on this group of hitters, missing the market's price by $7 per player. The only hitter he ties the market on here is Ellsbury.
The flip side of that is that he's kicking the market's ass on these hitters as a group. His prices told you to avoid Sizemore, Hamilton, Upton and Rios with a vengeance. In fact, the only hitter Rotoman is way below the market on with his projection and looks bad is Bay. This is a great set of projections.
Since projections include players who won't be on a typical auction roster, you can have more than $3,120 or $3,380 worth of stats. As a result, Rotoman's numbers needn't add up to these numbers. I am curious to see if he's "spending" more money in the middle than the market or not.
71-80 Most Expensive A.L. Hitters 2009
| # | Player | $ | Sal | +/- | AP | PK | 2008 |
| 71 | Lyle Overbay | $11 | $13 | -1 | $9 | $13 | $14 |
| 72 | J.D. Drew | $17 | $12 | +5 | $11 | $16 | $18 |
| 73 | Jose Guillen | $5 | $12 | -7 | $17 | $16 | $19 |
| 74 | Mike Lowell | $15 | $12 | +3 | $15 | $11 | $16 |
| 75 | Elvis Andrus | $18 | $12 | +6 | $3 | $12 | |
| 76 | Jason Bartlett | $28 | $12 | +17 | $8 | $12 | $16 |
| 77 | Matt Wieters | $9 | $12 | -3 | R1 | $14 | |
| 78 | Orlando Cabrera | $19 | $11 | +8 | $17 | $19 | $20 |
| 79 | Jack Cust | $14 | $11 | +3 | $11 | $17 | $16 |
| 80 | Aaron Hill | $27 | $11 | +16 | $11 | $8 | $4 |
| Average | $16 | $12 | +5 | $10 | $14 | $12 |
He most definitely is. Not only is Rotoman spending his money here, but he's kicking the shit out of both Patton and the market.
Putting aside individual player preferences, this is where Rotoman's pricing system speaks volume. If I could sum up what it is saying in a tidy few sentences, it would be this:
Players at the top are likely to fail. At least some of them are. Don't pay for peak performance. Guys in the middle will probably continue to linger in the middle, and will get underpaid by those paying $30+ hoping for continued $30+ performances. Pay the guys in the middle and hope that they stay there.To be sure, the one thing you don't want to do here is pay par prices for guys like Drew, Guillen, and Cust. These guys are profitable because the market discounts them. Take away the discount and you're buying a par team, whether you're using Rotoman's prices or the market's.
However, there's something to be said for Stage Three caution, and it appears that Rotoman is exercising that caution expertly.
Getting back to Alex...
It looks like he's trying to offer a compromise somewhere between Rotoman's conservative projections on the top guys and what the market is actually paying for those guys. The result is what you see above. Patton is coming closer to paying the guys at the top than Rotoman is, but he's still missing the mark.
What's intriguing is that maybe -after all of this talk of how Patton and Sports Weekly were underpaying - that Rotoman might be on to something.
If Stage One is starting to rear its ugly head, it may be time to go back to a good old fashioned Stage Three mentality.
5 comments:
Even if it's Peter making the projections and Alex doing the pricing, I'm sure Alex's projections wouldn't be that far off from Peter's.
And, I've found that Alex's pricing model values Batting Average especially and Steals more than other models. Perhaps it's how Alex builds the batting average formula compared to others.
I believe it was John Benson who built value from the lowest team batting average. This distributed the value throughout the league. Average was worth the same thing as HR, RBI, SB, and Runs. Alex's, of course, distributes value up from the league mean value and subtracts it from that same level down. Because it's a zero sum game the higher averages with higher at bat totals are worth more than in the other type of building value.
Alex's year end values take this into account. Although I do my own projections, I use Alex's formulae. I then bid based upon what the formulae spit out with my own style of inflation adjustment (which we've discussed previously). For that reason I frequently end up with Ichiro and last year found myself with Mauer. I think Alex doesn't bid quite what his own numbers put out for batting average. He seems to eyeball it a bit based upon expected price rather than expected value.
Recently, I've been doing the Patton projections. I think everyone who has paid the price knows this. I think everyone who has been following my projections since they were first published on ESPN.com in 1995 knows this. If they don't, whatever.
I didn't mean to sound so flip with the "whatever," I just wanted to be clear what Alex and I have been doing the past four years or so.
Another point: Neither Alex nor I base our prices on our projections. The projections are an educated guess, and they're helpful in characterizing what a player is expected to do, but pricing is really a matter of reading the market. Do you want the guy? And what is it going to take to get him?
If you always want him you make his bid price higher, if you want him some of the time you make his bid price par, and if you don't want him you give him a low bid price. This allows you to budget realistic prices for all players, and better manage your draft.
That's the idea, anyway.
I can't say for certain how long he's been at the wheel.
And I certainly didn't mean to sound like I have no idea what you guys are doing with that comment either. I literally meant I didn't know how long this arrangement has been in place, that's all.
I guess I see pricing as different and always assumed that you guys did as well. I never saw the bid prices as trying to figure out the market. I saw them as spending the $3120 among the players (ALx12) who should be drafted.
And, how would someone spend the $3120- to me it could be no other way than by what you think the player will earn. I may think the market will pay $42 for Ellsbury, but I only project him to earn $35 in salary. There's no way I put a bid price anything other than $35. If he goes more I rejoice. If he's going twice at considerably less I bid. If he's going twice at a little less I make a quick eval of my roster construction and money remaining and bid if it works. If he's on par, I just praise my projection as also matching the market.
While I redistribute some of the middle reliever dollars towards inflation (I've discussed my inflation theories before here), I don't cut salaries or raise them - otherwise the projections would be an extraneous exercise.
I don't mean to say that it can only be done my way- that stats projections equal dollar bids.
I know plenty of successful owners who just eyeball the bids based upon gut and then make them add up to $3120. Some successful owners don't even make the numbers match. They just know how to build a team and when to bid and when to stop. I've even known one good owner who just ranks the players by position and then just bids based upon the players around them once the market has been established.
It's very fluid, to say the least. I just think if you're going to go through the exercise of projecting hundreds of players it should be for more than entering the Fangraphs contest.
Post a Comment