Define ‘better’

The lead-up to Round 1 is a strange time. Recently I have caught myself about to make huge changes to my structure based on a rumour, swap out premiums because of a niggle, and second-guess my selections for any reason imaginable.

To be completely honest, I’m sick of it. I just want the season to start! I know I’ve put in the required research – like most coaches I’ve combed copious sources, stared at stats and pondered many opinions. My team will be as good as it can be.

But unfortunately I can’t let it slide. I sat down this morning and tried to decide who is a ‘better’ choice between two players and I just went round in circles. So I brought up DT Talk on my laptop and skimmed the comments – but it was just like seeing in writing exactly what the two voices in my head were arguing about.

What I needed was proof that I’ve made the right choice for me. Mathematical, statistical, definable proof – personal preferences and uniformed decisions are not going to help anyone win a car. And certainly not going to stop me stressing about trivial selections.

I needed a formula that could separate players – this way my bias would not be allowed to affect my decision. But in order to make a formula I needed to define what I wanted in a player, or in other words, the qualities I consider makes a player ‘better’.

So the trouble I was having was picking my 3^rd Midfielder. When it comes to Premiums, especially midfielders, my greatest concern (apart from average per game) is their consistency. I don’t want to be spending top-dollar on a player that I can’t rely on to pull through for me. Durability is another big factor – sure, Barlow is a gun, but I don’t have enough faith in his body anymore to be spending that much capital, just to sideways trade later on. Also, their ceiling and basement scoring is important to me for obvious reasons. So what I came up with was this:

I took the mean of their scores from last year and squared it. Squaring it gives more weighting to those with higher scores. Then I divided it by the Standard Deviation, which is a common measure of consistency. I then multiplied it by the fraction of the games they played in 2011 – i.e. by 1 if they played all games, 0.5 if they played half. I then multiplied this by the square root of the player’s Highest Score x their Lowest Score (square rooting helps limit its weighting). Then finally, I divided this Raw Score by 1000. Written out, this is:

[(Mean^2 / SD )* GamesFraction * sqrt(Highscore*Lowscore)] / 1000

I applied this formula to the top 15 averaging players from last season (ignoring Barlow), of which all are midfielders (Goddard, Chapman and Hodge had DPP).  I think it’s safe to assume the 3^rd best midfielder has a high chance of being apart of this bunch. These are the results:

GUN	2010 AVG	TBETTA FORMULA
Swan	123.1	86.97
Chapman	113.7	86.04
Ablett	119.0	80.58
Goddard	113.4	67.80
Boyd	115.2	60.16
Pendlebury	106.5	55.12
Hayes	108.0	53.16
Cross	103.3	52.90
Selwood	107.3	52.45
Bartel	105.4	50.15
Montagna	112.0	47.03
Judd	102.7	40.47*
Hodge	103.1	35.90
Dal Santo	103.0	32.30

*Judd missed three games through suspension last year, but missed no games through injury, so his score was calculated using a GamesFraction of 1.

This did not at all surprise me, because I had those top 6 players in my team at the end of last year! But there were a few things of interest…

• While Swan was easily the season’s best, Chapman only scored slightly less using my formula. This was mainly because of his super high consistency, and his lowest score was only 88 which didn’t hurt.
• Goddard also scored surprisingly well because of his impressive consistency (for the record, most consistent was Chapman, then Goddard, Ablett, Swan, Cross, Boyd).
• Montagna scored poorly due to his terrible consistency rating, and his low score of 67.
• Cross scored particularly impressively, considering his average was only 103.3. He also played every game and had a ceiling of 141 which helped his rating.

After this I decided to take into account price and therefore Value, so I divided the price by the Raw Score giving:

GUN	PRICE per TBETTA POINT
Chapman	$5.49
Swan	$5.87
Ablett	$6.14
Goddard	$6.95
Boyd	$7.95
Pendlebury	$8.02
Cross	$8.11
Hayes	$8.44
Selwood	$8.46
Bartel	$8.73
Montagna	$9.89
Judd	$10.54
Hodge	$11.90
Dal Santo	$13.24

This exercise helped me out a lot.  I recommend having a good think about what you prize most in a DTer and trying to quantify it somehow – it helps give you peace of mind. I’m confident in my choice now, and hopefully I can now get a good night’s sleep! It’s also important to note that this formula is my own personal take on what is ‘better’ for a Premium midfielder – I’m sure everyone’s take will be different.

Only problem is – now I want Chapman back in my team! Here we go again…

Let me know in the comments what you most value in a premium midfielder:

1. High Average
2. Consistent
3. Huge Scores
4. Durability

Cheers, tbetta