Transactions

Despite the restriction on transaction number and the resulting paranoia about saving some of those precious pick-ups for the final play-off run, we have still made 133 transactions through the first 8 weeks of this season.  For me personally, I like to fret and worry over a transaction before finally pulling the trigger in the morning, regretting it all day, and waiting to see if anyone picks up my playa off waivers taking advantage of my poor managerial decisions and over reactions.  I decided to look more closely at the transactions and see if there were any trends that I could pull out for our league.  I began by looking at the players that were picked up and dropped thus far. 

By and large, it seems like our teams are maintaining their makeups throughout the season. The positions with the greatest turnover are 2B and 3B with teams having 62.5% and 60% of those positions that they did at the beginning of the season.  LF and SS, however, have been the most added with 123% and 140% of the roster spots that they had at the beginning of the season.  With Placido Polanco having all sorts of hand problems, I have been in the market for a 2B and know that the position is quite sparse, so it makes sense that this position has been dropped so consistently, while the SSs that could fill in in the MI position have all been taken!  I was most surprised that the relief and starting pitchers have remained pretty consistent throughout the year.  Given that there will be at most 30 closers during the season, it makes sense that there wont be a mass exodus of closers.  The amount of moves for the RP and SP positions have been the highest of any position, though.  Also of note is that there have only been 9 waiver adds, which to me shows that there have only been 9 moves that others consider stupid enough to take advantage of.
Beyond the actual transactions, I wanted to see if the timing of our transactions had any trends.  I began by looking at which days we made transactions.  I went into this analysis with the thought that the most transactions were made on Sunday (to stream in those starters and get those extra Ks) and Monday (when the new week is starting and you can grab a 2-start starter or just get a fresh player).

While there were a lot of transactions made on Sunday and Monday, the most transactions were made on Wednesday and Thursday.  I suppose that by the time Thursday rolls around, you have a pretty good idea of how the match-up is going and you can make adjustments when you think they will actually have a real effect.  Next, I wanted to look at what times our transactions were made.

Aside from the 4:30am waiver-clearing time, it seems like the most transactions are made right around noon and then at around midnight.  Because we are only allowed to have a player play on the day he is picked up when the transaction goes through before the first pitch of the day (usually 1:05 or 2:20), it makes sense that we would want to scrunch those transactions in right before the deadline.  At night, after out players have let us down for yet another day, we like to go out and replace them in a highly spiteful and angry moment.
It would be interesting to see if these trends would hold up if we had unlimited transactions rather than capping then at 50.  I would assume that we would make many more impulsive transactions and the average time that a picked up player remained on a team would plummet, but I imagine that the trends regarding when and who is picked up would stay pretty much the same.  Thoughts?

For Your Consideration

Trends

When I think about my beloved team, The Ackbars, I often slip into the dream and actually believe that my team is a real team.  I find myself saying, "I am going to give Alex Rios a day off to rest that knee," or "I hate to drop Alberto Callaspo, he is such a great locker room guy."  The most frequent uttering to myself from myself in a world with only myself is, "My team is finally coming together and the guys are really clicking together."  Of course, these guys are all on real life teams and have nothing to do with eachother and, beyond the ten of us, are a completely boring agglomeration of playaz.  This week, however, I looked to see whether there was any truth to any of my notions that my team was "coming together" and if any of the categories were actually showing a trend upwards. 
I got the data together for each of the weeks for each of the teams and then performed a linear regression.  If a line that best fits the data had a positive slope, my thinking went, my team would generally be trending up in that category.  When I first did this, I actually did the graphs and, suffice to say, the "best fit" data was pretty far in left field when compared to the actual data point.  I got the R-squared for each of the data points and filtered the best fit lines through this measure of how well the line fits the data and only took those lines where r-squared is greater than 0.5 (on a 0 to 1 scale).  Those categories where, by this interpretation, the data had actual relevance are highlighted in nice calm blue.



The first thing that will stand out to you is that, really, there are very few trends for each of our teams.  The average for each week doesn't have any trends of significance, and even before the r-squared filter, the slopes of the lines were very low, showing that if there were any trend it would be very minor.
Second, I was heartened to see Omar's general offensive upward trend.  While I am happy for Omar and all of that, I was more pleased to see that, with his homers, his RBIs, avg, and OPS also were trending up, meaning that in general, this type of analysis works.  In the actual league, it appears that Omar is raking as of late, and his recent surge in the standings isn't simply the product of weaker opponents, as previous analyses have demonstrated. 
Conversely, I was disheartened to see that the Peachz'z pitching is generally staying pretty stable, with the exception of K/9.  The Peachz'z pitching has been dominant thus far into the season, and, unfortunately, it seems like it wasn't just a hot start. 
Other than that, my team the Ackbars, Woo Woo, and the Andrus Reclamation Project all seem to be pretty stable, which I know for my team is a mixed blessing; our hitting has been pretty good thus far, but the pitching has been very hit or miss, and it looks like it is going to be a long summer of hoping that with each start, the ERA and WHIP will go down, but after a few tenths, them just rising like my schlong when I see a picture of that Chelsea Clinton.  Hubba Hubba, Baby!

Batter than Average - UPDATED

One of my biggest issues with the H2H format is the so-called "Dallas Braden Effect," where your actual performance each week is not really important except as far as your single opponent's performance goes.  Rotisserie scoring takes care of this, but the problem is that after a certain amount of time, things will probably become pretty separated and the league will no longer be great to follow.  Moreover, being able to push the reset button at the end of the week is quite liberating; you need only worry about Trevor Hoffman's 6 earned run week for 7 days and then its gone like Keyser Soze.  I have thought about how things would look in a week-to-week roto league, and those results are interesting (and coming in the future), but not what I am looking at today.
No, today, we will examine what happens when your opponent for the week is the average of all of the other players.  This type of scoring is interesting because it actually changes the value of each metric; a home run a nice for you, but it also increases the average of homers, so it only has 90% value.  Conversely, when you hit a homer, you help yourself with that 90%, but also end up hurting not only your opponent, but all of them!  I can imagine how all over the country, people would be groaning when one of the numbers reached the critical threshold to turn over.  How sad would you all have been when my Dason Bay hit those two dingalings last night?

UPDATE - Something about the string of 0s and 1s was deeply unsettling to me, so as I was running, I decided that for anything to show up on this type of analysis, one week wouldn't be enough of a sample size.  Thus, I went back and applied this meathead to the entire seasons stats and the results were a bit more satisfying.



The most striking parts of this analysis is the overinflated record of Woo Woo and the underinflated record of Omar.  I have spoken previously about Omar and his horrible luck, but I thought that Woo Woo would have performed closer to his actual record than he did.  Also, I was surprised that Bernabe and Andrus could be any worse than their real record are, but alas and alak!
One of the things that I noticed about this analysis is that, unlike the real records, where each win must come at the price of a loss from another team, with the averages, you can have more wins than losses if there are a few bad apples who drag down the entire average.  In our league total, for example, we have a total record of 411-429 against the mean, which means that the mean was actually artificially inflated by a few awesome performances, and that there were more instances where we performed below the mean than above.
Once again, though, I feel like this would add a delicious twist to the overall scoring.  When you are ahead on homers, hitting another few and finishing 11-5 in that category doesn't really help you and, in fact, probably hurts you if you believe in the gambler's fallacy, which I think is more apt in baseball than other sports.  However, if this were the scoring system that we used, each of those additional dingalings would help to put your opponents farther and farther behind you by raising the overall average against which all teams are measured.

ESPN's Top-300s

As you may have noticed, ESPN recently released their Mid-May Re-Rankings for the top 300 playaz.  Seeing these sweet gooey numbers in front of me, I decided that I would see how each of the teams' power was stacking up.  Below, you will see listed the average power of each team in the league (the lower the better, like Pujols is a 1, Hanley is a 2, etc).  There are a few caveats to this analysis.  Notably, some of us have playaz on our teams who are not ranked in the top 300 by these ESPN fellows, in which case they were marked as "Weaklings" and assigned a value of 300 in the tabulation of power.  Furthermore, not all teams have the same number of playaz as some teams have DL spots occupied while others do not.  Because the power numbers were tabulated for the entire team, I figured that I should include everyone, active or not.
I was also interested in how the free agent wire figured into the teams' makeups, so I calculated how many playaz on each team was actually taken in the draft (whether for your team or for someone else's), and how many undrafted free agents populated each roster.
I find it interesting to note that, seemingly regardless of number of acquisitions, each team seems to have pretty much the same number of undrafted playaz on their teams (5-7).  I guess that this means that, once we find our good players, we stick with them and adopt a "last to come, first to go" type policy when picking up new guys.  Furthermore, each team has pretty much the same number of weaklings, showing that we, as a league, are doing pretty well in working that wire, as our commish says, "like a stripper pole."  Finally, it is interesting that these power numbers come fairly close to the standings, at least at the top and bottom.