As to raw bid for Sizemore, is simply the software (Patton$) with this league's denominators (and even with std. denoms his price is $28). Unless the software accounting for 10 teams is cheating players $.Ah. The price for Grady Sizemore's projected 2009 stats (24/70/27/.269) is $26 (not $28, though NSH might be in a 5x5 league), while Alex's bid was $38.
Neophytes to Patton's software don't understand how there could possibly be a $12 difference between Patton's projection and Patton's bid limit, and even experienced users often get tripped up.
For starters, the projections are no longer generated by Alex Patton. Rotoman has been partnering with Alex for a few years now, and the projections are his while the 4x4 bids are Patton's.
This is the primary reason for the discrepancy between some of the projections and some of the bids. Rotoman uses a method that seems to be based more on trends and traditional weighting than Patton's method, which was tied more to hunches and guesswork (though it certainly did consider past performance as well).
Looking at the bigger discrepancies in favor of Patton's bids paints a clearer picture of what I'm talking about.
| Rank | Player | AP | RM | '08 | '07 | '06 | RT$ |
| 1 | Grady Sizemore | $38 | $26 | $38 | $31 | $29 | $12 |
| 2 | Curtis Granderson | $27 | $18 | $22 | $32 | $14 | $20 |
| 3 | Aubrey Huff | $23 | $16 | $31 | $14 | $13 | $18 |
| 4 | B.J. Upton | $35 | $27 | $30 | $30 | $4 | $3 |
| 5 | Chris Davis | $25 | $19 | $14 | $6 | ||
| 6 | Mike Jacobs | $16 | $10 | $20 | $11 | $15 | $14 |
| 7 | Ian Kinsler | $34 | $28 | $33 | $22 | $16 | $34 |
| 8 | Alex Rodriguez | $27 | $21 | $37 | $53 | $33 | $12 |
| Average | $28 | $21 | $28 | $28 | $18 | $15 |
These are the eight hitters in the American League where Patton's bid (AP) was $6 or higher than Rotoman's (RM) projected earnings. And, so far, Rotoman looks like he's going to be much closer on these players than Patton is.
On the other hand, whose dollar amount do you like better?
I like Rotoman's Davis projection better, and probably would have fallen between Patton and Rotoman on Huff...had he been available in my league. For the rest of these players, I probably would have used Patton's bid or at least been much closer to it with my own bid.
We often think of robust projections when we remember that a projection is different than a bid. But this can work just as well with pessimistic projections.
Sizemore in his Age 26 season is going to lose $12 worth of value? I don't believe that, so I would have upped his bid considerably. The same principle would have applied for Kinsler, and to a lesser degree on Upton.
The Rotoman values are useful reality checks when you're looking at the Patton bids (or devising your own bids). But you must make sure to use projections as a starting point, and not the ending one.
2 comments:
Hi Mike,
Are you comparing Alex's bid prices with the prices my projections generate?
That's fertile ground for discussion, and I'm glad to see that my projection anticipated Grady Sizemore's struggles thus far, but my bid price for Sizemore (posted on The Guide's website on March 24th, was $38, just like Alex's.
Why would I regress Grady in my projection, but not so much in my bid price? The brutal answer is that our projections are at best rough calculations, and the fact is that a fairly large percent of what is going to happen in a baseball season for players is unknowable.
And another large percent is subject to random fluctuations.
Also, Alex's excellent software doesn't scale. So if my projections go over or under by some percentage, and they always do, the software doesn't adjust. I use the projection values as a double check, but the bid prices are what really matters.
My formulas for projection tend to rein in those who are galloping ahead swiftly, as Grady has been. But at the same time, while the math says Grady is a little more likely to disappoint than not, if he doesn't disappoint he will be a huge plus.
The projection reflects the odds, as my various tests reveal them. The bid price reflects the fact that you don't want him to go too cheaply.
I think next year Alex and I will offer up competing 4x4 bids, just for the sport of it. In most cases we'll be aligned, because of the market and how we understand it, but when we diverge you'll find out what we talk about when we're having a beer.
Best,
Peter
26 or 28, neither price was going to get Sizemore this year (even if Peter's projections turn out to be on the money). For our 5x5 AL only auction, with one team with an unsurmountable freeze list, I tweaked the software to get Sizemore to $30 and was willing to bet on 40% inflation. He went (but not to me) for $51.50 (and not tot he one team that should/could have bought him, the best freeze team). That team later acquired him for Chris Davis and junk (in an early, early fire sale). But for the inflation I see in a number of leagues the discrepancy between software and projection $ values would not have caught my eye. Having used the software for lo these many years, and swearing by it, I just need, I guess, to accept the inflationary nature of the game. Much like Paul Krugman advocated this morning I suppose, inflation is not the bogeyman at the moment.
Post a Comment