Keeping in mind that being an Auburn grad, I am very pro-SEC , let’s look at the sheer idiocy of this concept.

When matching two teams from the same conference with each other, you aren’t creating anything unique for the bowls. In fact, you are re-hashing a debate that has theoretically been settled. One team is the conference champion, the other is not. Do we really need another game that will bring the title of the conference champion into play? This is even more ridiculous when talking about conferences with either a full round-robin schedule or a Conference Championship.

Which bring us to the other issue, How can you be the National Champion when you aren’t even your Conference Champion?

Think on this for a minute. If you aren’t acknowledged as the best team in your conference, how can you be the best team in the nation? You might be the 2nd best team in the nation, but you should be out of consideration for number one.

I propose adding this simple rule:

Teams that are not recognized as their conference’s champion are not eligible for the National Title Game. Teams with no conference affiliation are eligible for the National Title Game.

The BCS title game should match the best two conference champions, maybe permitting the occasional independent into the game if they are good enough. It is already hard enough to accurately compare conferences without taking away this final test.

One last thought…people might THINK that the two best teams reside in the SEC, but if those teams don’t get one more chance to prove SEC “superiority” against outside competition, how can we be sure?

Remember, the true purpose of the National Championship game is to crown an undisputed number one. Having the SEC champion play another conference champion while the other top SEC teams play other opponents is the surest way to achieve that result.

This had been prompted by an online poll and I looked at the results of that poll in a second post. Feel free to look at it.  I’m focusing here on updating my rankings and sharing new thoughts.  I am including all relevant text and commentary from the original post in order try and keep key details together, like the methodology at the end of this post.

So let’s dive into the updated chart.  I used the old numbers and grabbed Ben Tate‘s stats from ESPN’s website.  I also didn’t drop Bostic even though he dropped out of the top 10 for comparison. Note that some numbers are different from before as I was able to get some better stats.

• Season stats refer to the best season
• Game stats refer to the best game
• APY: All-Purpose Yards
• RY: Rushing Yards
• TD: Touchdowns
• Records highlighted in bold.

Looking at Tate’s Career

Overall, a solid career, ranking 6th on the list.  He had the 4th best single rushing season and the 5th best career total rushing yards.  His all-purpose yards weren’t that strong, but neither was the offense during most of his tenure.

One thing jumps out at me, the high number of yards.  He clearly rushed for more than Stephen Davis.  What Tate is missing is touchdowns. Ben Tate was not a touchdown machine.  He averaged less than a touchdown a game in his best season. He is also missing a real signature game.  Sure, he hit 184 yards against Arkansas in 2009, but that was a loss and he didn’t hit the magic 200 barrier.

Tate does have the edge on Fullwood.  Their scoring is similar, but Tate has the edge in yardage.  Fullwood’s All-Purpose Yards pull close to Tate, but Tate clearly has the edge in career yards.

Just imagine if Tate hadn’t been in a dysfunctional offense for his sophomore and junior seasons.

My Methodology

There is a method in my madness. I took the categories above and ranked them in two ways. The first gave 10 points to the holder of the record, and the remaining got a fraction of that based on their score. For example, Carnell’s score for career rushing yards was 8.90.

The second gave points for top-ten finishes in each category, with the number one ranking getting 10. If someone ranked in the top ten more than once, they got more points. This led to Bo racking-up big points for top games as he as the 2nd, 3rd, 7th, and 8th biggest rushing games in Auburn history.

Also, while I didn’t account for this, let’s remember that today’s back have more games per season. While that may wear them down, it gives them more raw numbers.

Some season numbers may not be career highs for a person if their best season came when a better back led the way. If you find any of that, let me know and send me a link to the source.

Fun Fact: The record for most rushing yards in a game? Curtis Kuykendall ran for 307 yards at Miami on November 24, 1944. He did it on 33 attempts and got 4 touchdowns. It was World War II though, so the competition probably wasn’t all that impressive.

That being said, Jason Stark at ESPN wrote an interesting article the other week.  In his Rumblings and Grumblings column, Jason talked about Pushing for a minimum payroll threshold. In it, he discussed how teams get a LOT of money from the luxury tax on salaries.  He then suggested that their should be a penalty for teams that don’t spend a lot of money.

I had thought a lot about that previously and had come up with a system.  Essentially, I think that both the luxury tax, and the related reduction of that cap for “cheap” teams, should be based upon a more statistical measurement rather than some arbitrarily set maximum and minimum. By basing it on the number of standard deviations from the average (mean) salary, the amount would be market based.  This would make the levels change based upon the current spending levels.

The upshot is that if a team maintained a constant payroll, and the average increased, then they would pay less tax, or collect more tax.

Playing with Numbers

Using the numbers in Jason’s article, I determined that the average salary in 2009 was $95.8 million.  The standard deviation was $37.2 million.  So if you were between $58.6 and $133.0 million in salary, you were within one standard deviation.

That was broad, so I played with the numbers and came up with some rules:

• The bottom 15 spending teams could not be taxed.
• The top 15 spending teams were not eligible to receive funds.
• The tax starts at an eighth of a standard deviation above the mean ($100.4 mil) and increases at every additional eighth.
• The tax begins at 5% of total salary and grows by 2% plus an exponential component at every level.
• The tax payback starts at half a standard deviation below the mean ($77.2 mil) and increases at every additional fifth.
• The tax payback begins at 5% and grows by 5% until the first standard deviation and then increases by 10% from there. (5, 10, 15, 25, 35, 45…..)
• Rather than getting fancy, the money returned from teams spending too little are redistributed to all 30 teams evenly.  This only amounts to less than $2 million per team using my current calculations.

To help you get the idea, here are the two tax tables, skipping several rows for the tax levels so you can see the Yankee tax level. Notice that you can get to the point were you are paying more in taxes than you pay in salary if you go too high.

I was looking at a following article, and a reader had suggested that a winning team shouldn’t have to give back any of the received tax money.  That sounded very sensible.  Jason also mentioned not penalizing a team for the first drop, but I disagree.  It becomes complicated and if you don’t remove that exception, then only teams that are really doing it will try.

From those, I came up with some more rules:

• Teams with a .500 winning percentage or higher can keep all payments.
• Teams with a .450 winning percentage or lower have to give back the fully calculated enough.
• Teams between .450 and .500 are on a sliding scale, reducing the amount returned by 2% for every .001 winning percentage points above .450.

It is simpler than it sounds, especially ones you plug the formulas into a spreadsheet.

The Results

Well, here are the results.  The tax collected totals $348.7 million.  That is less than the $400 million Jason reports as collected. Here are the teams, salary, and taxes collected.

As you can see, the Yankees would pay a hefty chunk.  They are paying around $150 million now, and I tried not to increase their chunk too much.  I do believe that an exponential growth is required though.  That means that the 2nd $50 million over the average costs more than the 1st.  It should get more and more painful to increase payroll.

Now for the return.  I don’t have the formula that determines how the money is distributed, but here is the analysis of the teams that if they did receive funds, would have to return it.  The “Raw” values comes directly from the table above.

As you can see, the Marlins who would have had to give back 55% of what they collect can keep it all because they won 87 games.  The Pirates weren’t as successful and have to give back the 35% that they are docked.  The Padres were going to have to return 45%, but they managed a enough wins to reduce that by 26% and only have to return 33% of what they collected.

My Spreadsheet

I uploaded my spreadsheet to Google Docs, but it didn’t like all of my formulas.  It is there, but it isn’t as dynamic as it should be.  You can view it there and see all of the numbers and download it for your tinkering pleasure.  Change the payrolls of some of the teams and see how it impacts everyone’s totals.

If you want the raw Spreadsheet, I’ve uploaded it as a ZIP file, and renamed it with a .PDF so WordPress would be happy.

Let me know what you think.

I would like to now take that statement back.  For someone to be a “Challenger” this year, as defined in my last post, they would need to hold the top spot for ten weeks.  That isn’t going to happen.

The Numbers

I put together a spreadsheet that lets you enter the year and week for any point after the British Open.  Assuming Tiger has not played, it will show you his projected score for that week.  What does it show us?  Simple, in week 43 of this year, Tiger will have a score of 14.73.  Looking at the World Golf Rankings, that is five points higher than the current number two, Phil Mickelson at 9.64.

In fact, Tiger won’t drop below Phil’s score until Week 8 of 2009.  That is the end of February.  Tiger won’t drop below 10 until Valentines Day.

Now if Phil isn’t up to the job, there is still the matter of who will step up next.  Who knows?  There are still 5-6months of golf to play before we can expect someone to charge.  Don’t get me wrong, somebody can go on a tear and claim the top spot by year’s end. Tiger will be down to 11.66 in 2009, which is achievable.  It will take somebody stepping forward and not everybody taking turns winning.

Let’s see if the PGA Championship can clarify things at all.

Derek looks at several additional methods that were presented as feedback to his article, including my LOBR (R/LOB).  I was pretty pleased with that.  When he compared the results, the correlation of my method, 0.87, using data back to 1971 was highest of those using LOB (or BLOB).  The correlation of TOB to runs was much higher (called BRE in my last post), but that is to be expected (obvious statement #2 from the last post) and wasn’t the point of the original question.  All those numbers for 2007 are in my last post as well as in Derek’s 2nd article, just with different headers.

What is a Good LOB?

The original question in the article Stranded!, from BP reader JP, was:

Does the ‘Left on Base’ Statistic have any correlation to a team’s offensive success or failure?

My answer is simple.  YES!!!  However, just don’t look at the raw number.  Compare it to the Runs.  If the ratio, LOBR, is greater that 2:3 (R/LOB > .67) then your team did a decent job, win or lose.  If they perform higher than that over the course of the year, they’ll likely be in the top half in Runs Scored.

Looking at the All-Star game this year, the score was 4-3 for the American League.  However, the LOBR wasn’t pretty for either team.  The AL had a LOBR of 0.235 and the NL had a LOBR of 0.273.  The NL had the better LOBR, but it was horrible and only a slight advantage.

In a more day-to-day measure, the Braves just played a three game set to the Nationals.  Here are their LOBR stats (LOB).  Guess which games they won.

• Braves 1.400 to Nats 0.857
• 0.250 to 1.000
• 0.545 to 1.250

If you said that the Braves won the first game and then lost the next two to the Nats, you are correct.  In the first game, the LOB is pretty close, as was the score, 7-6.  In the second game, the LOB was the same, but when you look at the LOBR, you can tell that the score wasn’t close.  It wasn’t as it was 8-2.  The third one finished with a nastier score of 15-6.

These games are a perfect example of looking at LOB in context.  By itself, it doesn’t mean much.  When you use Runs, you can then determine if your team is failing to get those runs around.  The Nationals had a higher LOB in two of the games, the big blowout and the one that they lost.  In the smaller blowout, both teams had the same number of runs.  The Nationals were more efficient (1.074 LOBR for the Nats vs 0.625 for the Braves) in the series and won it.

More runners is still key and will always give a better correlation, especially over time.  LOBR will give you instant feedback on how effective your offense was firing on any given day, or for the whole season.

Which brings us to the big question, who is the next number one? Tiger still holds the spot, even after not playing for a month. The results of the The Open Championship are unlikely to change the leader board, except to bring Tiger closer to the pack.

If you look at the current standings, Phil Mickelson is best positioned to take the spot. I have my doubts about his ability to pull it off. He isn’t doing well at the British this weekend. He cleared the cut by two strokes, but was eight strokes back with a +8. He then hit 6 over the last two days and should finish well out of it.

After Phil, it is wide open. Whoever wants it more can take it, so it will be interesting to see.

A Historical Look

Grabbing numbers of weeks in the top spot, I’ve thrown a little graphic showing the Leaders (most weeks at #1), Challengers (10+ weeks or 2 times at #1), and Notables (at least one week at #1) per year. It isn’t meant to show the best player each year, but to give an indication of dominance in the golfing world.

As you can see, Greg Norman and Tiger Woods have dominated the charts since the rankings came out in 1986. Vijay Singh would have a stronger showing except his time at the top was over two years and I selected calendar years. Vijay is the victim of the methodology that I chose. Many would argue that he was the best player of 1994, and I would probably have to agree.

The chart really expresses dominance. Tiger I knew. Greg Norman also dominated, which I thought, but didn’t understand the scope. Nick Faldo had a decent run for a few years, but he didn’t define an era. He was just a break from Greg.

Tiger already has this year wrapped-up. The question is whether anyone can perform well enough to hit the challenger level, or will they be left at Notable. I expect at least one person to hit the Challenger status this year and maybe even a new Champion for next year. We could see a fair amount of turmoil for a while as it will probably take Tiger at least 6-7 months to challenge for the top spot, most likely longer.

I can’t wait to see who will actually pull this out.

[Edit 2008-07-21: Looked at the updated ratings after The Open Championship. No real changes except that Padraig Harrington has moved into 3rd place, noticeably in front of the pack. Tiger still has 2x Phil's score in 2nd and 3x over Harrington.]

The Obvious

Let’s cover the obvious. The more people that you put on base, Total on Base (TOB), the more people will score. It is a simple concept that powers Batting Average (AVG) and On Base Percentage (OBP). Derek ran some numbers for about 918 teams over the course of 25+ seasons and sure enough, there was a strong correlation.

Next bit of logic, the more people that you put on base, the more people that you leave there. Very simple. If you, on average, got 40% of your base runners home, then you would have a higher BLOB with 2000 base runners than with 1900 base runners. The numbers confirm the correlation, not as strong as with runs, but strong.

Finally, he compared the number of Runs with BLOB. This also had a strong correlation, but it was the weakest of the three. This may seem to indicate that a high BLOB is good. Well, over the course of a season it isn’t bad as it indicates more men on base which leads to more Runs.

Derek then concludes that saying a high BLOB, while frustrating, may be a good thing. This is where I pick it up.

Base Running Efficiency

I decided to check the ratio of Runs to BLOB in 2007. Derek used numbers from 2007 for his quick illustration, and I’m going to use the same numbers, mostly because I have them readily accessible from Derek’s article.

The more Runs a team scored, the higher the ratio. What is very interesting was that the ratio for the majors last year was .666. That means for every two LOB, a team scored a Run. Those that fell below that ratio, faired poorly. The top 15 teams all had ratios higher than .666. Of the bottom 15, only the Blue Jays ranked at 17, had a score above the mark. They came in at .677, but were in the bottom 25% in TOB. Back to obvious statement number 2, more on base, more runs. Let’s call this the Left on Base Ratio (LOBR)

To look at it another way, the lower scoring teams got 33-35% of their base runners around to score. The higher scoring teams, around 40%. The number 15 team, the Tampa Bay Rays, got 37.3% of their runners across. Their ratio of Runs to Everyone that scored more than them (756 Runs) had a better ratio of getting those runners home. Only 1 of the remaining 15 teams below them had a better rate. Let’s call this Base Runner Efficiency (BRE).

Here is a quick chart with all the numbers for 2007. Totals for the majors are at the bottom.

One thing that you realize, and can enter the obvious discussion, the more people a team can get on base, the better they are at getting them home. The very act of getting on base can push existing runners closer to home, or the rest of the way there.

What can we do with this in evaluating our daily frustration? Let’s look at a recent slate of games apply the norms for the majors to the games.

Any Given Day

Here a a chart showing the Teams, Runs, BLOB, LOBR, and Result for Sunday, May 18 2008 I’ve paired the teams from each game together, Winner and then Loser.

Go National League! 8-6 on Sunday. What we are looking at though is the LOBR. Here is the average for the Winners and Losers:

• Winner LOBR: 0.935
• Loser LOBR: 0.543

The Winning teams did leave an average of one extra runner on base. However, they scored twice as many runs, winning by an average score of 7 to 3. It was a good day for all the teams as the overall ratio for the day was 0.754.

A quick note of exceptional games:

• Boston over Milwaukee (11-7 Runs, 9-2 BLOB). The Brewers were very efficient in getting their people home. However, the Red Sox got them by volume and were still over the 0.666 average at 1.222.
• St. Louis over Tampa Bay (5-4, 16-5) Another case of sheer volume. Tampa did a great job, they just needed more.
• Houston over Texas (5-4, 9-3) Just like the St. Louis game, just smaller LOBR numbers. Two one-run games. Reinforces the importance of getting on people to start the process.

The other 12 games went to the winner in LOBR. If I had a full database, I’d pull more games in and see how it all maps out. In any case, this can serve as a starting point in understanding the role of BLOB in greater detail.