CVCR Question

[viper]

I am trying CVCR variations on weighting -specifically comparing players values with all OF, CA/OF mix and normal position mix.

After running the first two sets of dollar value I discovered that the value assigned to Runs and RBI slightly changes with both RBIs and Runs having a lower dollar value on the 1.2/0.8 scenario. The incremental value per run/RNI is the same [or damn close] so it has the be the replacement level. Is this totally caused by replacement players when HRs having a higher value and SBs being downgraded? This would make sense as HRs are essentially gone once replacement are determined by SBs are not.

I ran a NFBC type roster setup using .280 as my baseline average and .70 as my hitting percentage of the $260

[Todd Zola]

It is not the replacement level, that is fixed. It is almost definitely a difference in pool size factor as that changes the marginal amount assigned to each category to size the pool.

[viper]

I noticed that it changed. I wonder why would the factor change?

[Todd Zola]

This is explained in the essay. Because with the differing weights, the number of players with $1 or greater value has changed.

This is due to the fixed nature of the replacement. The older, more labor intensive, hard to automate method used the replacement level to size the pool. Something has to be tweaked to to fix the pool size. The CVRC tweaks via a marginal pricing component. The "by-hand" method tweaks via the actual replacement level.

[viper]

sounds reasonable to me. I'm trying to run six scenarios and see how players value is changed. I'm mixing Hr/Sb at 1/1 and 1.2/0.8 against players as all UT, CA/OF and by MVPOS. Interesting results.

Oh, and I have been lured to that dark side by Lord Zola and am leaving any SGP calculating. I just need to find the mix of category value that best suits my beliefs in category value.

[cwk1963]

sounds reasonable to me. I'm trying to run six scenarios and see how players value is changed. I'm mixing Hr/Sb at 1/1 and 1.2/0.8 against players as all UT, CA/OF and by MVPOS. Interesting results.

Oh, and I have been lured to that dark side by Lord Zola and am leaving any SGP calculating. I just need to find the mix of category value that best suits my beliefs in category value.

You mean lured from the dark side. Welcome to the light; the light is your friend.

[viper]

I have some differences with some of the replacement assumptions used but this approach is actually better than my old approach. Plus, I can easily generate a multitude of variation by using CVCR - something I couldn't do before. I can simplify my spreadsheet but it will have to wait until after all my drafting is completed. Too much is inter-twined and I have put together a semi-automated means to get data prepped for movements into the two CVCR packages and into my own specialized tiers.

[shif6]

I ran a six pools for an 8 team al. I ran off the values, but there were not 16 positively valued catchers. Are there supposed to be? If so, what am I likely doing wrong?

[Todd Zola]

This is explained in the CVRC tutorial, after using the pool size factor, a few catchers can be knocked down to $0 or -$1. Those catchers should just be considered as $1.

As an aside, I strongly suggest using only 2 pools, C and non-C as the multiple eligibility of so many players is such that you get a better idea of relative value.