Welcome back to my run-down of each division in baseball. Over the past six weeks I’ve bored you with a look at the junior circuit. Although a little out of date, my synopses of the AL East, Central and West are darn good reads if I say so myself!
Now it is time to turn to our attention to the senior circuit kicking off with the NL East. Over the next month we’ll also look at the rest of the National League—rest assured no team will be left out.
In 2006, the Braves’ run of 14 consecutive division titles (bar 1994) came to an abrupt halt as the Mets marched comfortably to their first postseason appearance since the 2000 subway series. In other 2006 division news, the Phillies fell a tad short of the wild card again, and the Braves were marooned, surprisingly, with a losing record.
Pundits and analysts have had their fingers burnt many times trying to forecast the NL East. Many a naysayer perennially predicted the end of the Braves’ run but with little joy. Ironically, last year was probably the first year where the common media view turned from “they won’t win” to “I’ll always pick the Braves until they don’t win”. With the latter having happened that meant that 2007 predictions were always going to be more objective.
To prove the point here is what the THT staff said in our preseason poll:
1st 2nd 3rd 4th 5th Richard Barbieri NYM PHI ATL FLA WAS Sal Baxamusa PHI NYM ATL FLA WAS John Beamer ATL NYM PHI FLA WAS Brian Borawski PHI NYM ATL FLA WAS John Brattain PHI NYM ATL FLA WAS Matthew Carruth PHI ATL NYM FLA WAS Chris Constancio NYM PHI ATL FLA WAS David Gassko NYM ATL PHI FLA WAS Ben Jacobs PHI ATL NYM FLA WAS Larry Mahnken PHI ATL NYM FLA WAS Dave Studeman ATL NYM PHI FLA WAS Steve Treder NYM PHI ATL FLA WAS Bryan Tsao NYM PHI ATL FLA WAS John Walsh PHI NYM ATL FLA WAS Geoff Young NYM ATL PHI FLA WAS Consensus PHI NYM ATL FLA WAS
One of us has to be right surely!
If we take a look at the various projection systems we see similar ambiguity.
CHONE Diamond Mind PECOTA ZiPS W L W L W L W L New York 86 76 83 79 85 77 87 75 Philadelphia 86 76 85 77 88 74 87 75 Atlanta 83 79 85 77 81 81 87 75 Florida 75 87 73 89 77 85 72 90 Washington 69 93 75 87 66 96 70 92
I couldn’t quite believe that no projection had the Mets ahead. The uncertainty is confirmed as only one, PECTOA, didn’t predict a tie division and none expects any team to win more than 90 games.
What’s clear is that New York, Philadelphia and Atlanta are all very much in the hunt. Florida are not expected to repeat their mercurial 2006 year and Washington, well, they are doomed to anchor the division and very likely to end up with the worst record in baseball.
Current State of the West
Here are the NL East standings as of Sunday, July 8 (before Sunday’s games):
National League East Team W L W% GB Pwin Diff New York 48 38 0.558 0 48 0 Atlanta 46 42 0.523 3 47 1 Philadelphia 43 44 0.494 5.5 44 1 Florida 42 46 0.477 7 41 -1 Washington 35 52 0.402 13.5 39 4
As advertised it is all fairly close at the summit. And here is the story pictorially, courtesy of the legendary THT graph:
The story is relatively simple. The Braves and Mets both started explosively and opened up a double digit lead over the Phillies and Marlins by mid-April—indeed after opening 1-9 it looked as though the Phillies were out of the equation altogether.
However, regression to the mean swiftly kicked-in as the Phillies went on a tear while the Braves and Mets both slipped up, so the division is the three-way gang-bang predicted. The Fish are also on the fringes but the losing record and young, inexperienced team will likely see them ebb away. To the surprise of no one, not least the Washington glitterati, the Nationals are all but out of it.
Interestingly none of the top four teams deviate much from their pythag, indicating luck hasn’t been a huge factor—but trust me, that picture was dramatically different a month or so ago with the Braves and Mets overperforming and the Phillies underperforming. The Nationals are an astonishing four games over their expected record—without that they’d be a sub .400 team. Ouch.
Let’s see what the THT statistical model spits out when asked who will triumph in East given each team’s start.
To recap the methodology, what we do is use THT projections to calculate a team’s expected wins above replacement. This is based on player depth charts that I pulled together for each ball club. The division win probability is calculated from win distribution curves that work out the odds of each team finishing on top given their talent—for the technically minded, I use recursive probability functions. Here are the results (assuming that current standings reflect true talent):
30th March 2007 8th July 2007 W L Playoff% W L Playoff% New York 85 77 36% 90 72 51% Atlanta 82 80 24% 84 78 23% Philadelphia 81 81 20% 81 81 14% Florida 80 82 17% 80 82 12% Washington 70 92 3% 66 96 1%
Compared to some of the other projections we saw earlier, the THT numbers look in remarkably good shape. The order is correct and the number of games each team is expected to win hasn’t changed much either—especially when you consider that the standard deviation of talent projections is around 10 wins.
The New York Mets have played a bit better than expected, largely due to an overachieving rotation, and should get better when Pedro Martinez dons his uniform in a few weeks’ time. However, unlike last year no team is dominating the division. Atlanta, Philadelphia and Florida are all expected to finish within two games of where we predicted. No doubt each team will now go on a horrendous winning or losing streak and make the predictions look utterly foolish.
To all intents and purposes the THT model is saying that there is a 90% chance that the winner will come from the top three teams—shush, don’t tell Vegas!
The Wisdom of the Crowds
Who says the crowd is stupid? Along with the THT projections, the Tradesports prediction market was one of the few bodies of opinion that has, so far, called the division correctly—with the Mets winning the division, counter to preseason conventional wisdom. Okay, so you can grumble that it was hotter on the Phils than Atlanta but given their respective records over the past couple of months you’d probably side with that view.
Here are the win percentages now and back in late March:
Team Market Win % (March 31st) Market Win % (July 8th) New York 47% 70% Atlanta 16% 15% Philadelphia 28% 13% Florida 8% 2% Washington 2% 0%
Let’s step through each team’s win expectancy graph as the season has unfolded. First we’ll take the two northern giants, New York and Philadelphia.
As the season dawned the Mets were an even bet to take the East. After an explosive start that saw them win their first four games by outscoring opponents 31-3, the market readjusted its win expectancy to 60%. From that point through to early June the Mets machine continued to roll forward on full power as they amassed a 34-18 record and were comfortably the best team in the NL. That record was built on pitching as Glavine, Maine, Perez, Sosa and Hernandez all recorded sub-4.00 ERAs. However, since then the wheels have come off slightly with a 12-19 record and they have been sucked closer to both the Phillies and Braves.
The Phillies’ story is almost antithetical as they opened up with 10 losses in their first 13 games and saw their win expectancy plumb from 30% to 15%—once more the Phillies contrived to throw their season away in the opening weeks. However, since that nadir, regression to the mean and then some kicked in and the Phillies have gone 42-30, including sweeps against their two division rivals.
Win expectancy isn’t just influenced by your team’s performance but also by your opponents’ performance. This is seen when in mid-June the Phillies’ win expectancy leapt from 10% to around 25%—this was not only because the Phillies were playing well but also because the Braves and Mets were falling back towards earth. Phillies fans must be hitting their collective head against a brick wall … had the Phillies gotten off to a semi-reasonable start they would have had a healthy division lead at that point.
And now the southern giants, Atlanta and Florida:
The Braves’ story is similar to that of the Mets with a white-hot start followed by some decidedly mediocre baseball. They opened with seven wins in their first eight and finished April with an impressive 16-9 record. Another good run in early May saw their record stretch to 24-12 and with it the promise of a new dynasty, perhaps. Perhaps not—the market certainly didn’t think so as the Braves’ win expectancy remained stubbornly below the Mets’ despite a similar record.
However, that was as good as it got and the Braves reverted to the .500 team that they probably are. June 2006 was a horrid month and 2007 was almost a repeat as the Braves had a shocking interleague schedule playing 15 games against four of the best teams in baseball (Boston twice, Detroit, Cleveland and Minnesota). In that stretch they slumped 4-11 and the malaise had set in. By late June they were back to .500 before a mini-resurgence has seen them claw back some ground from the Mets.
The Marlins’ story doesn’t quite have the same romance as it had last year. Expectations were for no more than a .500 team and, so far, they haven’t exceeded that and have managed to keep their head firmly below the radar on the way—they have not been particularly streaky, have never really threatened the top three, and have also played significantly better than the Nats, who are marooned in last place. The Fish’s win expectancy has naturally eroded simply because they have been slowly bleeding out of contention.
The Nats’ win expectancy never really got off the runway and a lack of liquidity means that the graphic isn’t particularly meaningful.
PROTRADE is a sports stock market where you can buy or sell players and teams depending on whether you think they are over- or under-valued based on a fantasy scoring system.
As I discussed in a previous column this is less useful for players because arbitrary statistics are used, but for teams the scoring system is more sensible so we can build a picture as to how well people think each will do.
Here is a reminder of the points scoring system:
Price Criteria $1 for each regular season win $10 for qualifying for the playoffs $4 for each playoff win $10 for winning the league division series $20 for winning the pennant $30 for winning the World Series
Have a look at how the price of the NL East contenders has fluctuated since mid-April (when PROTRADE first launched this feature).
Team Price (April 21)Price (July 8) Earnings (July 8) P/E New York $131.98 $134.00 $47.00 2.9 Atlanta $104.06 $108.27 $45.00 2.4 Philadelphia $96.55 $105.19 $44.00 2.4 Florida $82.21 $91.06 $41.00 2.2 Washington $60.32 $62.90 $35.00 1.8
As things stand the crowd appears wiser than the analysts and pundits, with the Mets well pegged to win the division from day 1 and still very much in the ascendancy. One has to be careful with PROTRADE data because the market is based on play money rather than proper cash. This means that fans can sometimes back the more popular teams because they have less skin in the game.
As far as PROTRADE is concerned the Mets are practically guaranteed to make the postseason, and possibly the NLCS. The Phillies and Braves, with a price of between $105-110, are very much in the playoff mix locking horns with various NL West teams for that last coveted October slot.
The price/earnings ratio doesn’t tell us too much in this case except that New York are expected to outperform in the second half, which is something we can blatantly see just by looking at the prices!
Before I started penning this column I thought the NL East would be the most exciting division in the bigs because of the closeness of the teams. While that may be true on the field, when it comes to looking at the numbers there is not much of controversy to report—it’s all a bit dull really.
References & Resources
Merci Tradesports and PROTRADE for providing a wonderful set of data.