Announcement

**RAY@HQ** · 03-14-2005, 08:07 PM

Quick answers:

1. a difference of one or a couple of SB can have a big impact.

2. you also have to look at the number of projected AB... the BA projection is worth more if it's based on more ABs.

I believe those two items explain all of your examples.

**BrooklynMonster** · 03-14-2005, 08:30 PM

Umm ...not really

Konerko's 1 projected stolen base and +3 rbi is worth MORE than Thome's additional 18 runs 4 homers and 4 pts of batting average? That's just wrong.

I respect that Bonds fewer at bats lessens the impact of his BA but his average is projected nearly 60 points higher than B. Giles 4 more runs 14 more homers 11 fewer rbi and 3 fewer sb's yet Bonds is valued $4.36 less?

Plus Sheff and Matsui projectons are nearly identical with Sheff having one MORE sb but he is valued $3.84 less

Thanks for the quick response but I just don't think that cuts it. Does anyone else feel these results are off or is it just me?

**RAY@HQ** · 03-14-2005, 08:44 PM

Brooklyn,

Took a closer look... I admit that I am leery of the Thome/Konerko example. I'm going to look more closely at that one.

As I said earlier, the Sheffield/Matsui difference is that Matsui gets ~70 extra ABs at that .291 BA. Is that worth $3? Perhaps.

There have been anecdotal reports of the weighting of BA breaking down with strange/partial AB total. That could be what's going on with Bonds.... the combo of very high BA and few ABs is extreme, and may not be getting weighted properly.

Thanks for the feedback, I'll pass it along.

**RotoMaster** · 03-14-2005, 11:36 PM

The results I get from the custom valuator (NL 4x4 10-tm $260) are so out of whack that I've decided I can no longer use them for my draft. I'll just use the standard R$ and adjust accordingly using my own deflation techniques.

It's too bad. I've been a long-time subscriber and really liked the idea of the custom valuator, but I just can't trust it. Hopefully it'll be fixed by next year.

**RAY@HQ** · 03-15-2005, 07:13 AM

Rotomaster, please give me something more to go on than the results are "out of whack". Don't know whether you saw it, but we had a very lengthy thread a week or two ago, where users pointed out issues with the CV tool and we got several of them addressed. The tool's by no means perfect, but I've looked at it extensively over the last few weeks and have repeatedly concluded that the values are good enough for use.

If you don't mind, post your specific inputs and which values you think are "out of whack", and I'll take a closer look and (if necessary) get the issues passed to our tech team, even if they're only on the list for next year.

**RotoMaster** · 03-15-2005, 04:06 PM

Thanks for the feedback Ray.

There are too many examples to list them all, but here's a few:

(10-tm NL 4x4 $260 14H 9P)

GENERAL
- For starters the top 230 players does not add up to $2600.
- The 90th pitcher should be $1, yet from the 82nd pitcher on all the values are negative.
- The 140th hitter is $0.01 (which is wrong but acceptable), but the 141th hitter plummets to -$5.50. This screws things up when you have to switch a few guys in-and-out of the top 140 (due to position requirements).
- Top hitters like Pujols, etc. should be worth less in a 10-team league vs. a 12/13 team league, yet if anything it's the opposite.

PROJECTIONS - There are endless ones here that don't make sense but here's a couple:
- Compare Casey's projections to Bonds. Now explain to me how Casey is $32 in the custom valuator and Bonds is $22.
- In the standard values Abreu and Helton are worth the same. When I switch to 10-team custom values Abreu stays the same but Helton goes up by $5. Why would Helton go up while Abreu stayed the same? Helton's strength is average while Abreu's is speed. So this tells me that in my 10-team Helton's "extra" average is more valuable than Abreu's "extra' speed. This doesn't make sense. When moving from a 12 team to a 10 team, average should become less valuable not more (ie more total team ABs, better "average" player, etc.)
- There are a bunch of others but I'm tired of typing.

Look I'm a happy long-time subscriber and I'm not trying to be a pr**k here. The fact is there are many quirks and I just can't trust the thing. Plus if the total values won't add up to $2600 then what's the point in the first place. I'm still gonna have to use some deflation techinques to get it to the proper $2600 starting point.

I think HQ should tear the thing down and rebuild it from scratch.

**RAY@HQ** · 03-15-2005, 04:35 PM

Rotomaster, just about all of the things you bring up are known issues. See the "Custom Valuator - Screwy Results" thread for details, they're on the "to do list" already. Couple of other points.

- total values may not exactly equal $2600, as the programmer's idea is to give different options for the last roster slot depending on what you need. I ran a 10-team last night and got $2640 for total value, which for my purposes is acceptable. Your mileage may vary, but that's a pretty small % error.

- the last value being $0.01 instead of $1 is on the to-fix list. I don't quite understand how the negative values work, I've already asked for explanation on that. But the FAQ explains why we don't move the $1 level to support position requirements.

- I believe I determined yesterday that Bonds' combo of very high BA and very low AB total is being incorrectly valued. Awaiting confirmation of that. If there is a BA-weighting issue, it may also affect the Abreu-Helton comparison. Again, though, I'm not sure I agree with your blanket statement that SB should be worth more than BA in the shallower format. Fact is, again as noted in the FAQs, that the projection files are built on the SGP method and the CV tool built on Scarcity. That will lead to fluctuations in value even if you compare identical formats. Is one more right than the other? Depends on your purpose.

- I disagree with your assessment of what should happen to dollar values of top-end players in shallower leagues, I don't think the rule you suggest is necessarily true.

Again, I do appreciate the feedback. Sounds like you've run into a lot of the same issues that I've seen from other users in the last few weeks. The sad/annoying fact of it is that issues like the $0.01/$1 valuation are very noticeable and undermine the credibility of the tool, and on the usability of the tool, but they don't undermine it's core function.

Just as you're not trying to be a pr*ck, I'm not trying to blow off your concerns. We know there's still work to be done with the tool, and we're getting it done. We made a major push around the first week of March to deal with the biggest issues, and as a result the tool is a heck of a lot better now than it was a few weeks ago. We're not in the clear yet, but I'm urging people not to just ignore the tool, it is a very powerful part of our arsenal and the results are "good enough" to convey that value. That said, we're still pushing to make it better. But as Ron has said in other threads on this topic, this is crunch time for us and there has to be prioritization of where tech resources go during this period. Thanks for sticking with us.

**Doc** · 03-15-2005, 06:49 PM

I certainly hope these grids are accurate. My auction was last sunday. we do both an nl only and an al only 12 teams each with different catagories and different salary amounts. Just to give some ideas of the "bargins" that I hope I got.

2nd base 30-35 grid. Giles $38, Kent $29, I got vidro for $20
Starting pitching 25-30 grid. prior 41, clemens 30, peavy 32, perez 27, other perez 19, I got Thomson for 9
Starting pitching 20-25 grid. hudson 38, zambrano 34,beckett 30, burnett 25, smoltz 21, I got weaver and petittie for 15 each.

I don't even want to go into the AL. I have always found the grid to be the best auction tool ever to quickly compare players with relativly the same "value" and see where the bargins are. But, this assumes that the "relative values" are correct. Kind of a big IF this year. But, I am keeping my fingers crossed. thanks for the great efforts ray and tech guys. doc

**RotoMaster** · 03-15-2005, 08:58 PM

Thanks Ray.

My suggestion would be for HQ to rebuild its own custom valuator using the same SGP method they use in their standard projections (for consistency). You have all the formulas to convert SGPs into R$, I don't see why it would be so hard to adapt them for different league sizes and rules. I do understand though that obtaining the SGPs for different league sizes with different categories is tedious if not impossible. But I also think SGPs could be simulated without league history data.

Regardless, the BEST solution for me would be if you gave us the formulas for converting SGPs into R$, then I could just plug in my league's historical SGP data and get my own projections.

Anyways, I'm done griping. The only reason I gripe in the first place is cuz I know HQ will listen and try their best to improve their product. Thanks again.

**PHIL@HQ** · 03-16-2005, 11:57 AM

Rotomaster, you are so right. We are listening and are trying to make things perfect. In the meantime, I agree with Ray that as is the Valuator is still a useful tool.

**derossi** · 03-24-2005, 05:06 PM

I will be VERY disappointed if the value calculator isn't trustworthy soon.

Originally posted by RAY@HQ
The sad/annoying fact of it is that issues like the $0.01/$1 valuation are very noticeable and undermine the credibility of the tool,

BINGO!! The huge negative numbers and the $0.01/$1 valuation issue tell me something IMPORTANT is broken. That scares me.

I consider that calculator ESSENTAIL, and will be very disappointed if it isn't fixed next week.

BTW I accept the slight bust in total values (roundoff error) and the lack of catchers with positive value (because it is true!).

I have written my own spreadsheets in the past. I think I can differentiate the symptoms of a minor issue from those of a major problem. I am afraid of that calculator right now.

I am on my knees! Please get it fixed! Thanks.

**RAY@HQ** · 03-24-2005, 05:14 PM

Was there a point to bumping that quote without comment, derossi?

To be perfectly clear, the context of the quote was me saying that the tool is very useful, but some obvious yet not serious flaws like the $0.01/$1 thing make people think the tool is more flawed than it is. Kind of like not buying a house because the front door's off the hinges, but everything inside is beautiful.

Again, the tool's not perfect yet... but you get my point, I hope.

**derossi** · 03-24-2005, 05:22 PM

Sorry. My previous post has been edited to be useful.

Apologies. I hit the wrong button while typing and posted before I was done writing. OOPS!! :-)

I was editing the post (to make it useful) while you responded.

**RAY@HQ** · 03-24-2005, 06:07 PM

Fair enough, derossi. Didn't mean to jump on you, it definitely seems that we're on the same page. And I'll say it one more time: while we're still working on the tool, I wouldn't hesitate to use it 'as-is'.

Announcement

Custom Valuator Problems

Custom Valuator Problems

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment