Wednesday, May 25, 2005

dog ages

Here's a graphic showing how old DoG has been through its history. It's ordered by present age. Gray cells are 33+, and yellow cells are 25-, as of some point in the season.

For me, it's interesting how many guys younger than I am have retired.

If you want to play "Survivor", here's who left the island (that had been on since the first episode) each year
1995: Seeger, Jethro, Dennis, Joel, Mike, Scott
1996: John Bar, Gary
1997: Justin
1998: no one
1999: Cork, Bob
2000: Lenny, Bickford (leaving with perfect 46-0 records with DoG at Nationals)
2001: Mooney, Coop
2002: Jordan
2003: no one
2004: no one (leaving Billy, Alex, and me left)

Here is the same information sorted kinda by seniority:


Which do you like better? Which conveys the flow of the team better?

And here's one that shows the +/- 1 sigma ages as well as the max and min.
A box and whiskers plot might be better, but I think I need an add-on in Excel to do that.

7 comments:

Marshall said...

In this case the first one appears to show the flow of the team a little better to me, because the start and end point teams are a little better grouped. There may be a way to show the overall flow of the team's age distribution that would be more effective at a single glance in some graphic form, though it wouldn't have all the supporting data.

Anonymous said...

so was 98 the oldest and smallest team?

luke

parinella said...

Oldest but not smallest. I added the average age and the number of players and updated the graph. Smallest was 19 the first two years, then 20 the next three, then we hurt ourselves for the next five years by winning with 25 in 1999.

Anonymous said...

I'm a little confused. Are you suggesting that by carrying 25 guys in '99, the younger half of DoG did not have the advantage of extensive PT to benefit them in the later years (2000 to present day)? How did carrying 25 guys hurt?


- Lenny Lime

parinella said...

No, I'm saying something different. I think 25 is too many to have, but since we won in 1999 with 25, it showed that it's ok to go with the large roster. Had we lost with 25, after 5 years of winning with 19-20, after 2 years before that of losing with 25, that would probably have convinced us that 25 is too many and something like 20 is optimal roster size. It really hurt us in 2003 when we had 28 playing.

Anonymous said...

To put it simply, why do you feel that 19-20 is the optimal roster size? (I know this is off-topic, but I'm rather interested in the notion of roster limits and the like.)

parinella said...

There is no such thing as "off-topic" here.

Before I start, let me say that I strongly believe that roster size should be up to the team to decide, not limited by the UPA. It may be that in 5 or 10 years that it's appropriate, but not now. We're supposed to let individuals decide for themselves.

You need a big team in order to have good practices and to make it through Nationals, but you want a small team so that your good players play just the right amount and so that the lesser players play an adequate amount. The underlying assumption is that there is a noticeable dropoff in ability as you go down the roster, the extent of which of course depends on the team and its environment (specifically, does the 21st player have other worthwhile options?).

19-20 players gives you 2-3 subs per squad at practice if everyone is playing. I think 3-4 would be preferable, but 4-5 is too many because you don't get to play enough.

So why not 21-22 then? Let's start by looking at the top of the roster. I think that the limit for how much a fairly well-conditioned athlete should play is close to 50% of the points (maybe it'll be 60% for an extremely well-conditioned one who is not injury prone). This applies to men's Nationals. A good Nationals team might have six players who fit this bill. That takes up 3 spots out of 7, leaving 4 spots for the remaining players. How much do you have to play to get in the rhythm of the game and be effective? Let's say that it's 25%, so that would mean that we'd have an additional 12 players, bringing our roster to 18.

So, our ideal tournament roster is 18, and our ideal practice roster is 22. 19-20 is a compromise leaning towards the tournament.

If the sport becomes more mercenary or more professional, or if the team is happy with players 21-N getting no playing time at Nationals other than garbage time, then I think the roster can be bigger. And of course you need to have contingencies.