Value Differences 12/14 vs 12/21 Valuator

[rbbl]

I ran dollar values for my league(AL, 5x5, 9 tm, Top Man) on 12/15 and 12/21. The values for SP went down and those for RP went up. What caused this? Offense values stayed the same.
Examples: 12/15: J. Santana 9.85; 12/21 9.48;
M. Rivera 6.96; 8.17
Bedard 6.67; 5.67
Papelbon 6.74; 7.85
Sabathia 6.21; 5.77

Thanks, Patrick

[Michael@HQ]

Patrick, you might want to check to see if the total number of saves in the AL projection population went down sharply from the 12/14 projections to the 12/17 projections. That might do it.

[rbbl]

Michael

Good thought. However, there is only a difference of 3 saves between the 2 data bases.

I remember that the values seemed to change last year from week to week. Values for individual players should change as PT, IP and other stats are adjusted as we get closer to opening day. But, I am suspicious when a general adjustment takes place between starters & relievers.

Hopefully, someone from BBHQ will respond and enlighten us.

Patrick

[Michael@HQ]

My thought was that if a guy's individual dollar amount changes but his projection stayed roughly the same, then there must have been a change in the rest of the player pool that made his stats comparatively more valuable. Didn't work here though.

I'm sure someone from BBHQ will speak up -- especially if you and I keep bumping the thread! It might not happen until the tech support guys' next business day, whenever that is.

[RAY@HQ]

Not to dismiss the concern, but I'm not sure 5 examples constitue a global change. Santana and Sabathia's changes wouldn't even be noticeable if not for the fractional dollar values.

These changes are almost certainly due to wider issues in the playing pool. I'll be honest, I don't see any shifts in those values above that are really concerning. As Michael noted, you'll see more of this "wiggling" as we get closer to Opening Day.

[rbbl]

Ray, I delayed this response even though I had more than 5 examples. In fact, I calculated the $ differences for every pitcher from the 12/15 and 12/21 data bases. I was hoping this was a one time problem.

However, the 1/5 db values using the valuator again have made huge swings between SP and RP, mainly closers. Something is going on here.

J. Santana values went from 7.36(12/15); 7.10(12/21); 7.24(1/5).(-3.5% to + 2%).
Putz: 5.60; 6.49; 5.56( +15.9% to -14.3%)
Rivera: 5.22; 6.31; 5.34(+20.9% to -15.4%).

I can go on with examples(again I have compared the values for every pitcher in the data base) but take my word for it RPs'/closer values have gone up and then back down in large %; the SP values have gone in the oposite directions for smaller %s but I guess that is because the $ are spread over a lot more SP vs Closers.

The underlying stats for the SP and closer have not really changed over the last few projections. In fact, I checked the 5x5 $ values of the 1/5 vs 12/21 db and there is little difference between them. This means the varying values between SP and closers is being caused by the valuator which is based on my league variables(which were the same for each set of values).

Unless I can find out why the valuator keeps changing its calculations, I don't know which set of values are valid.

I would appreciate your help in resolving this.

Also, I am wondering if anyone else out there is experiencing this problem?

Thanks

[RobR]

If you hadn't said 1/5, but said 1/7 or 1/8, I'd believe this is due to a bug I introduced on 12/16 when I rolled out category weights support and corrected yesterday on 1/7. The bug severely devalued pitching K's in 5x5 leagues. Since top relievers usually suffer and top starters usually benefit from K in 5x5 the behavior ties out, except for the date.

Are you sure of the 1/5 date? If 1/5 is accurate, would you be willing to see what difference you see between 1/5 and 1/8 data?

Could you please post the full parameters you are inputting? From your dollar values, I'd guess it's not a $260 league. Are you using the default roster position allocations? Are you using any other non-default options?

[rbbl]

RobR: Valuation Dates/Changes/values

RobR,

The 3 valuation reports I ran were not 1/5, 12/21 and 12/26( I used those dates because I thought the valuations were based on those projection files) but the valuation reports were actually run on 1/7, 12/15, 12/28.

If a bug involving Ks was introduced on 12/16, it showed up on my 12/28 report. My 1/7 values were similiar to those in my 12/15 report and supports your suggestion that the fix of 1/7 worked. Bottom line, the closers are worth less and the SP more which makes sense if Ks were involved. It now appears the values of the 1/7 report are accurate.

I just ran the report again, 1/8, and a quick review indicates the pitcher values are the same as the 1/7 report!

My specs are 9 team, al, $65, 70%off, 0 bench, use fractional $, force positions, Top Players, Rep. Sal $0.25, P 11, OF 4, C 2, 1B 1, 2B 1, SS 1, 3B 1, MI 1, CI 1, UT 3(actually we have 1 DH and 2 UT), no scarce, 5x5.

Thanks for your help.