Well, I don't know if this has already been posted, but I found an article by Tristan Cockroft at ESPN talking about how starting 2 start SP's is not a good thing to do. I just had to disagree (of course, it was because the leagues where I have weird scoring rules are affected)...here was my reply.
Mercer Boy wrote:Just skimmed your article on the two-start SP strategy. I just wanted to comment on it.
While it may not be effective in a normal H2H 5x5 league or even in a roto league that uses ERA and WHIP as categories and daily roster changes, it is absolutely vital to do it in weekly H2H leagues that use IP as a category (yes, I know that's not common, but I am playing in one). If you don't pick up at least one 2 start guy, the chances of you winning IP and K is about nil.
It's also a must to do in weekly points league that give high value to wins and IP. Last week I lost my H2H points matchup by 70 points because my opponent had 9 total starts to my 6 (I had very good pitchers - Mulder, Vazquez, Garland, Mussina, Oswalt), but only one pitched twice. He won 7 of 9 starts, and I won 5 of my 6. He also had 20 more IP than I did. These extra stats (not including the points he got for K's) gave him an extra 80 total points than my pitchers. My offense consists of guys like Pujols, A-Rod, Abreu, Jeter, Bay, Alou, and Hafner (no, I'm not making this up) - all who had big weeks - and I still couldn't even come close because of the points he scored from his SP's.
Sure, if you take into account all 2 start pitchers, their combined stats may be very bad...but that's why you only pick certain guys each week. All you have to do is study the matchups (who's playing two bad teams, who has a good track record against the two teams they play, etc.) and determine which guys have the best chance to do well twice.
So as you can see, in some instances and scoring formats, you absolutely have to use 2 start SP's. I have at least one if not two revolving 2 start slots on my fantasy teams in these leagues because it's impossible to win without it.
this guy's gone off the deep end. if you look at his two charts at the bottom of the page, they say that two start pitchers have more than twice the expected win total, twice the expected K total, and better ERA and WHIP than 1-start pitchers... what a crock!
his analysis is completely off. why does this guy have a job?
"I don't buy everything I read,
I havn't even read everything I've bought"
"I find it more comforting to believe that this [life] isn't simply a test."
Joined: 16 Apr 2004
Home Cafe: Baseball
Location: at Morimoto's, eating $50 worth of sushi
he really doesn't know what he is talking about -- that much is clear. His analysis is always poor. He and Skip Bayless should have a kid together. I think Karabell is the best one they have.
Of course, it didn't seem like he really gave it his due as a head to head strategy. Tell me until the cows come home, but there is and obviosu truth to the fact that if you play only one game you need to win 100 percent to get ONE win. Two, you only need to win 50 percent. Factor in k's will definitely be higher -- you realize you are potentially losing era and whip.
He suddenly wants to revist common strategy based on a small smaple size and a poor understanding of the numbers he is reading.
Well, I'll say this--if the guy thinks you get the probability of getting a win in two starts by adding the chance of getting a win in one start to itself then he is a statistical idiot!
Note for anyone who cares: the proper way to use the 34.4 percent win statistic in this case is to reverse it (i.e, the probability of not getting a win in a given start is 63.6) and raise it to the second power. Thus, the probability of not getting a win is 40.4 percent and conversely the probability of getting at least one win is 59.6 percent. (You'll note that this is indeed very close to his data, which shows that 58.5 % got at least one win.)
Don't have time to drill through the whole thing, but this is such a basic error it calls into question the rest of his analysis.
Wow. It's amazing they pay this guy to do his job. What are his qualifications again? I agree on the point about the chart. That chart clearly illustrates why using 2-start pitchers are a good idea on every level. And with that data, he argues the other side? Simply amazing.