Week 11, The Best of the Best

 

 

Cory just got back from a UK golf trip so he sent along some photos with his pops and Royal Porthcawl. No they didn’t win the Ryder Cup, the course had just hosted two Ryder Cups so the course made a fake one to make you feel cool. It’d work on me.

2024 YTD Power Ranks
TOTAL			HITTING			PITCHING
1	Arthur	4.00	1	Max	2.38	1	Dean	3.50
2	Michael	4.29	2	Arthur	4.13	2	Michael	3.83
3	Dean	4.50	3	Paul	4.38	3	Arthur	3.83
4	Max	4.57	4	Brian/Josh	4.50	4	Robert	4.17
5	Brian/Josh	5.36	5	Michael	4.63	5	Keith	4.50
6	Robert	5.93	6	Dean	5.25	6	Cory	5.33
7	Paul	6.07	7	Dave	6.38	7	Dave	6.00
8	Dave	6.21	8	Robert	7.25	8	Brian/Josh	6.50
9	Cory	6.57	9	Cory	7.50	9	Max	7.50
10	Keith	6.64	10	Keith	8.25	10	Paul	8.33

 

I decided to let ChatGPT handle this week’s power ranks analysis, looking at the year to date chart:

1. Overall Trends:

   - Arthur (Orange): Arthur started strong and maintained the top position for most of the weeks but showed a slight decline towards the end.

   - Michael (Blue): Michael has shown a steady performance, improving slightly over the weeks.

   - Max (Green): Max showed significant improvement over the weeks, especially after week 5, reaching a high rank towards the end.

   - Dean (Navy Blue): Dean started low but has been improving gradually, particularly in the last few weeks.

2. Volatility:

   - Robert (Yellow): Robert showed a significant drop after the first week and then fluctuated but showed improvement in the latter part of the period.

   - Cory (Brown): Cory's rank has been quite volatile, showing both significant drops and gains throughout the weeks.

   - Brian/Josh (Dark Yellow): This team experienced a steady decline after the initial few weeks and maintained a lower rank towards the end.

3. Consistency:

   - Paul (Gray): Paul has had a consistent performance, staying in the middle range without significant fluctuations.

   - Keith (Light Blue): Keith also showed a relatively steady performance, with slight ups and downs but remaining in the middle ranks.

   - Dave (Light Gray): Dave's rank has been relatively stable, with minor fluctuations around the mid to lower range.

4. Notable Improvements and Declines:

   - Dean and Max: Both showed notable improvements over time, with Max showing the most significant rise in ranks towards the end.

   - Brian/Josh and Cory: These teams experienced the most notable declines, particularly after initial higher ranks for Cory.

In summary, Arthur maintained a strong position overall, while Max and Dean showed significant improvements. On the other hand, Brian/Josh and Cory experienced notable declines, indicating struggles over the weeks.

Couldn’t have said it better, myself. Thanks AI.

OK cool, but HOW good and HOW bad have those weeks and runs been. I have a new stat for you.

I had a thought recently about how we can compare our performance over time: week over week or year over year.

I’ve started to run self-generated ranking systems for players before each season. The idea behind it is that Player A may be awesome at home runs and project to hit 50 of them, but the league average in home runs for a player is only 25, so that player puts up TWICE as many home runs as the average, so he gets a value of 2.0 for that category. BUT he may only hit .220 while league average is .260 so he’s 15% worse than league average so he gets a -0.15 value for AVG. Add up all the values and you get the player values projected for the season. The same concept can be applied to our whole team fantasy week to week outputs. We can compare week to week outputs against league averages to come up with a valuation for that week.

With the stats I keep I have all of our week to week outputs by fantasy manager going back to 2012. This year, the ‘average’ fantasy week (across all ten teams and all 11 weeks) looks like this:

R	HR	RBI	SB	BB	K	AVG	SLG	NW	K	ERA	WHIP	QS	SV+H
34.6	9.1	33.5	6.1	25.8	60.6	0.247	0.412	0.6	76.0	3.642	1.202	5.3	5.4

So when Dean puts up a week like he did in week 11, you can see why he probably had a lot of positive values in a lot of the categories

R	HR	RBI	SB	BB	K	AVG	SLG	NW	K	ERA	WHIP	QS	SV+H
50	20	55	4	19	68	0.304	0.587	5	72	4.24	1.03	4	7

 

There is one more bit of math to this that involves dividing the difference between the output and the average by the standard deviation (this helps normalize values like the ratio categories where the numerical difference is tiny (0.260-0.220=0.40 in the previous example) compared to the counting category numerical different (50-25=25) in the previous example). If you care about the math from here, let me know. I’ll skip to the next part. Here is how valuable Dean’s week 11 was

R	HR	RBI	SB	BB	K	AVG	SLG	NW	K	ERA	WHIP	QS	SV+H	VALUE
2.2	3.4	2.8	(0.7)	(1.1)	0.7	2.2	2.9	1.3	(0.2)	0.6	(1.0)	(0.6)	0.8	13.3

Conversely, here is what a less than awesome week looks like, here is Cory’s week 9:

	R	HR	RBI	SB	BB	K	AVG	SLG	NW	K	ERA	WHIP	QS	SV+H
Outputs	32	4	23	2	23	71	0.201	0.33	3	58	2.14	0.92	6	5
Values	(0.4)	(1.6)	(1.3)	(1.4)	(0.4)	1.0	(1.7)	(1.4)	0.7	(1.0)	(1.6)	(1.7)	0.3	(0.2)

Interestingly, Michael played Dean in week 11 and Cory in week 9.  

A few things. First, how well do averages of week ranks correlate with our year to date power ranks?

Manager	Average of Week Ranks through week 11	Total Power Rank through week 11
Arthur	3.66	4.00
Michael	2.99	4.29
Dean	2.88	4.50
Brian/Josh	2.46	5.36
Paul	2.22	6.07
Max	-0.36	4.57
Dave	-2.02	6.21
Robert	-2.29	5.93
Cory	-3.54	6.57
Keith	-4.06	6.64

 

Pretty close, though not exactly equivalent. What this tells me is that Brian and Josh have had higher week to week fluctuation (higher highs than lower lows) which has kept their week to week average up but let their power rank sag a little bit. Meanwhile Max is the opposite, his week to week output has lower lows than higher highs (see the previous blog for reasons why) leading to being a below average team more often than not week to week, but overall his power rank is strong because his weekly lows don’t pull down his power rank very much relative to how low his pitching power rank already is due to his low overall volume. We know who his team is and a poor pitching week just doesn’t harm his power rank very much...but it does harm his week to week average rank.

Next, what have been the best weeks so far this year?

Noting the decimal places for values that should be whole numbers like Ks or SB, for the 10 or 11 day fantasy weeks I divide the totals by 1.5 to standardize week vs week analysis. For example Robert had 5 NW over 10 days, 5/1.5=3.33. Two weeks each from Arthur, Michael, and Brian/Josh here. Each have had a handful of excellent weeks.

Of course now we have to look at the worst weeks:

Cory’s had a touch run here since week 4. It’s ok, his golf game is crushing it right now though. Keith also makes it three times on this chart, but ChatGPT seemed to not hate Keith’s outlook, so we’ll stick with that.

All this is to say is that I have a new tool now to evaluate week to week performance. No more blogs of saying “[Person XX] had an OK week as they beat [Person YY].” I can now say “[Person AA] put up a 1.50 week average as opposed to competitor [Person BB] who only did a -2.0,” and who doesn’t love putting numbers to feelings…amirite…

One last table for funsies, the 10 all-time best weeks (remember these stat averages reset every year, what made an awesome week in 2019 is different than what makes an awesome week now):

Yes, Brian and Josh have the best week all time, I expect this to be in their LinkedIn profiles by the weekend.