Projections - skills versus playing time

[viper]

It seems this board is slowing up some so I have resorted to glancing at other boards. Embedded in a Shandler site thread, I came up with what seems to be interesting topic while reading one of Todd's comments.

With early projections coming out soon, I am wondering how valuable they are before we answer numerous questions concerning who is playing where. You may be able to project a player's skills at this early date [subject somewhat to the lineup around them and the park they play in] but for many players, playing time is a huge guess. This leads to my question.

Which is more critical in player projections, skill projections or playing time projections?

[alleyoops]

I'd rate them as equally important, as if either is off by a substantial amount, the projection may lead to a poor result in strategy related to that player.

If I had to choose between the two, I'd pick skills, mainly because it is easier to make the appropriate adjustments to the numbers when playing time changes (i.e. it's just math). If skills change (Delmon Young, anyone?), it's really tough to know how far to move which numbers - at least for us laymen.

[cwk1963]

I say skills since if they are there the playing time should come. Conversely, playing time will/should drop if the skills aren't there. Of course there's the outliers such as Jose Offerman getting playing time over Justin Morneau .

[da_big_kid_94]

If that precept is being broached as a general rule of thumb, I can see where it may have some merit. If it's being put out as some sort of all encompassing ideal, then I would take it with a huge grain of salt... in fact, I would cite current instances where it is NOT the norm but the exception. CWK just so happened to mention of of the teams that this is a prime example of. There's another run by the current AL manager of the year. Any coincidence that they ran 1-2 in this years balloting?

[Guest]

The most critical thing is playing time.

Skills are skills, give me a three year average and you're going to get somewhere close. This is why the MARCELS do so well as a general projection methodology.

We sprinkle some magic fairy dust on our projections and I think that makes them very good, I think last years projections demonstrated a good basis for our methodology. This year we are tightening it up even more.

But ultimately at the end of the year you told me you would either give me the actual PA/IP for every player in the majors or their skill rates (HR/AB, K/9, etc) I wouldn't even hesitate to ask for playing time.

Now, the real question ensuing from this should be "has anyone demonstrated a real capacity for projecting playing time"? I would say the jury is out on that one.

[Todd Zola]

To piggy-back a little on Gary's comment and to tease a likely new feature to our projections, a means of "customizing" them.

We are contemplating coming up with a means for you to select a set of skills that best suits your philosophy. We will soon explain, but there is wriggle room in certain aspects of projection methodology. In theory, you could select the skills translation you like best. These projections will be presented as if EVERYONE gets the same at bats or innings. This way you can compare them on a strict skills basis. You can then say "give me a set of projections using this skill set with your playing time projections" or you can take the time to enter your own at bats/innings. Of course, you can do what you always have been able to do and tweak our playing time projections to your liking and input then into the CVRCs.