Pot Boost Probabilities

[DevilsGM]

This seems the best place for it. Coming up to the draft POT boosting prospects will be an obvious tactic. Is anyone aware of and willing to share the general probabilities involved with POT boosting?

[SharksGM]

I intended to test it at some point but never got around to it. If you're motivated, try making a file just before with rollover with all U20* boosters of various POT/CON and let us know.

*Players over 20 have very small odds of boosting even with CON below 75 but I'm not sure how that works.

[BLUESGM]

I have boosted Carlsson and Farrance thus far - Farrance boosted right away - Carlsson - played him as an underager in AHL and he boosted the next season. I didnt do extensive testing to be able to say what probability is.

[DevilsGM]

I thought boosts took a year to take effect (the next 21.6 switchover)

[Jets GM]

No, that is not true. Used to think that but sign and send back to juniors after drafting before the rollover gives you a chance.

I have found celebrating when a player boosts, and whining about when one does not, increases the likelihood in future.

[DevilsGM]

I have found celebrating when a player boosts, and whining about when one does not, increases the likelihood in future.

Lol, noted

[JetsGM]

*Players over 20 have very small odds of boosting even with CON below 75 but I'm not sure how that works.

20 to 25 yrs old - con 70 or higher - pot 59 or lower

I think the under 20 year old pot boosters is pretty much general knowledge at this point... under 20, pot 69 or lower, con 75 or higher, can boost 10/20/30 pot points at once or multiple times throughout their youth up to 7X, higher con seems to convey more chance of boosts, but I never got hardcore enough to quantitate the qualitative on this.

Best bet is to run some test sims before the rollover, I used to export to excel before rollover, sim past rollover then export and run a spreadsheet comparison, you get a real good idea of whats what and seeing as ur a stats guy NJ, you can probably just "adjust" 10 or 20 guys to whatever variables you want and after a few test sims gather up some decent data.

One crazy anomaly I found though is older guys going up by exactly 1 pot point... essentially changing their character role by having this happen. Noticed it mainly in older D... was super rare happening once or twice in like ten or so test sims... would be cool if you saw this being replicated as I thought it was kind of interesting and out of the realm of common EHM knowledge.

Or you can just do what I do... ask Vik if you have an EHM question... I believe he has achieved EHM god status with all his contributions and understanding of the game.

[DevilsGM]

10 to 30 edits? What is that a sample size for haha. I plan to run 2 tests. Rub the default file multiple times through the rollover to see general frequency. Then rewrite the entire roster file with 18-38 year olds with specified cons and pots.

[JetsGM]

10 to 30 edits? What is that a sample size for haha. I plan to run 2 tests. Rub the default file multiple times through the rollover to see general frequency. Then rewrite the entire roster file with 18-38 year olds with specified cons and pots.

Lol, so let me guess your current career is either engineering or accounting.

Cool to see some results on a mass scale man, most interested in any hidden "anomalies". Used to love doing this stuff with EHM... until I realized I would end up being a crappy GM no matter how much I knew about the inner workings.

[SharksGM]

Rub the default file multiple times through the rollover

Whoa, too much information there.

[SharksGM]

On a more serious note, if anyone has some data on whether player roles actually matter for anything, let me know. I've been assuming that they don't.

[DevilsGM]

I tried looking at player role and how it effected development briefly awhile ago. I didn't really see anything of note. The 79 guys "Character Player" seemed to overdevelop a bit mroe but the sample size I got after taking out all the noise was very small and consisted of only like 5 guys so nothing definitive by any means.

[DevilsGM]

Just finished one season and took a look at the boosts. 24 in total. 3 were aged over 20 so looks like thats a lie, think it still has to be ELC though. The POT drop limit is typicially advertised as below 70 POT and below 70 CON, but I saw POTs from 50 -89 and the CONs were all below 50.

[SharksGM]

Huh? Pot drops only happen below 50 CON. Where did you read otherwise? Also, they have to be signed.

[DevilsGM]

Don't recall. COuld have been a random forum in one of my many searches, could have been simply word of mouth.

[Vik (Habs)]

I think it's <70 CON in 1.18.

[Dallas Stars GM]

When will I know if some of my players boosted or not?

[DevilsGM]

The boosting update happens on the 20th when all the stats and everything are updated. Pot drops can happen throughout the season though.

[SharksGM]

I finally got around to doing this test. I filled the file with players born on the 1st of the month from 1998/06 to 2000/12 (yes I know Dec. 2000 birthdays aren't even draft eligible but it didn't end up mattering).

- For '99 and '00 born players with CON >=75, there is a 45% chance of boosting.
- The chance of boosting does not depend on POT or CON.

- For '98 born players, only those under 20 years of age boosted (i.e. 1998/06/01 birthdays did not boost at any CON).
- I only did 10 players per CON for '98 born, but only 21% boosted overall.
- None of the '98 born with 85 <= CON <= 89 boosted at all.

So basically if you picked a '00 born CON booster, they have a 70% chance of boosting this year or next.

[DevilsGM]

I've been playing a single player game in 1.18 and looking at the restuls on the side, and things actually differ quite a bit.

- Age range requirement appears to be repalced with a "Rookie" tag requirement.
- Older players still boost less often likely pointing to age based probabilities for boosting.
- POT had no effect within the boosting bounds
- CON had an effect, players in the lower range boosted less often, and it appeared to pick up greatly at a CON of 80.

So in 1.18 if you had a 17-year-old draftee who had a 90+ CON and you signed him he was more or less guaranteed to boost.



On the other side of things, one interesting thing I noticed is that pot drops happen all the time during the regular season. Not sure if that is something that was in 1.16, but might be something to keep an eye out for.

[Vik (Habs)]

I finally got around to doing this test. I filled the file with players born on the 1st of the month from 1998/06 to 2000/12 (yes I know Dec. 2000 birthdays aren't even draft eligible but it didn't end up mattering).

- For '99 and '00 born players with CON >=75, there is a 45% chance of boosting.
- The chance of boosting does not depend on POT or CON.

- For '98 born players, only those under 20 years of age boosted (i.e. 1998/06/01 birthdays did not boost at any CON).
- I only did 10 players per CON for '98 born, but only 21% boosted overall.
- None of the '98 born with 85 <= CON <= 89 boosted at all.

So basically if you picked a '00 born CON booster, they have a 70% chance of boosting this year or next.

Are you factoring in that only signed players can boost? Because I just ran a couple tests and 45% doesn't seem like it can be right. Also, CON definitely plays a factor (higher is better).

[SharksGM]

Are you factoring in that only signed players can boost? Because I just ran a couple tests and 45% doesn't seem like it can be right. Also, CON definitely plays a factor (higher is better).

Every player was signed and on a pro team on the 20th. I had 100 players each with every CON from 75 to 99 and there was no dependence on CON at all.

[Vik (Habs)]

Are you factoring in that only signed players can boost? Because I just ran a couple tests and 45% doesn't seem like it can be right. Also, CON definitely plays a factor (higher is better).

Every player was signed and on a pro team on the 20th. I had 100 players each with every CON from 75 to 99 and there was no dependence on CON at all.

I did one test where every player had 78 CON and another where every player had 99 CON. The first test had ~75% boost, the second ~96% (1362 players eligible for the boost each time, all with the same birthday).

[SharksGM]

Try it for yourself then: http://ehechockey.com/boosttest/players.ehm

You can use that with the current file after advancing to the 20th.

I tried it again and here are my results, pretty much the same as before: http://ehechockey.com/boosttest/players.ehm

For 1999 born:

Code: Select all

CON     #    boost      frac
75	50	24	0.492
76	50	26	
77	50	21	
78	50	25	
79	50	27	
80	50	21	0.42
81	50	22	
82	50	19	
83	50	27	
84	50	16	
85	50	21	0.424
86	50	18	
87	50	22	
88	50	22	
89	50	23	
90	50	26	0.476
91	50	19	
92	50	28	
93	50	26	
94	50	20	
95	50	18	0.448
96	50	25	
97	50	20	
98	50	23	
99	50	26	
			
	1250	565	0.452

2000:

Code: Select all

CON     #    boost      frac
75	50	24	0.412
76	50	22	
77	50	20	
78	50	16	
79	50	21	
80	50	21	0.416
81	50	16	
82	50	25	
83	50	22	
84	50	20	
85	50	19	0.44
86	50	22	
87	50	25	
88	50	21	
89	50	23	
90	50	23	0.52
91	50	35	
92	50	26	
93	50	26	
94	50	20	
95	50	22	0.472
96	50	24	
97	50	25	
98	50	24	
99	50	23	
			
	1250	565	0.452

[Vik (Habs)]

You don't have 50 players eligible for boost for each CON. I think you're counting both unsigned players and goalies. For 1999, I ran it with your file and got the following:

Code: Select all

CON	#	Boosts	%
75	40	25	0.625
76	30	22	0.733
77	31	25	0.806
78	32	19	0.594
79	33	21	0.636
80	26	15	0.577
81	26	24	0.923
82	27	24	0.889
83	37	29	0.784
84	19	13	0.684
85	27	22	0.815
86	20	15	0.750
87	25	21	0.840
88	31	27	0.871
89	31	28	0.903
90	29	25	0.862
91	26	21	0.808
92	31	26	0.839
93	31	26	0.839
94	22	18	0.818
95	20	18	0.900
96	26	23	0.885
97	22	21	0.955
98	24	21	0.875
99	26	25	0.962
			
All	692	554	0.801

[SharksGM]

You raise a good point. Everyone was signed, but I forgot to change their teams to pro teams from 98/99. I wonder how the probability of a player being on a pro team changed with row number to conspire to make the pot boost chance 45% across the board.

Anyway, having fixed that, the odds of a boost are pretty close to CON%. That's even higher than I thought. But if it's really as high as 70% for a 75 CON prospect, then the average 75 CON draftee should only fail to boost twice 9% of the time. I could swear that the fraction of failed single boosters is more than 10%.

[Vik (Habs)]

Anyway, having fixed that, the odds of a boost are pretty close to CON%. That's even higher than I thought. But if it's really as high as 70% for a 75 CON prospect, then the average 75 CON draftee should only fail to boost twice 9% of the time. I could swear that the fraction of failed single boosters is more than 10%.

I tried it out last year and it was a big reason I figured I wasn't entirely crazy to go after Kostin (if anyone hadn't figured out why my initial tests were with 78 CON).

[DevilsGM]

All the pot boosters. GG San Jose on the Godly Gildon.

Code: Select all

Name			POT	CON	Age	POT2	Pot Change	Team
Logan Hutsko		58	77	19.5	78	20	NJ
Alexis Binner		54	75	19.7	74	20	TB
Maxwell Gildon		68	86	19.2	78	10	SJ
Timothy Liljegren	69	80	19.3	79	10	NYR
Ilya Karpukhin		68	75	20.1	78	10	ANA
Klim Kostin		67	78	19.3	77	10	MTL
K'Andre Miller		66	78	18.5	76	10	VAN
Ryan Merkley		69	78	18	79	10	SJ
Jonatan Berggren	68	79	18.1	78	10	LA
Kirill Marchenko	66	76	18	76	10	CGY
Alexander Khovanov	64	75	18.3	74	10	MTL
Semyon Der-Arguchintsev	68	76	17.9	78	10	SJ
Jack Perbix		68	75	17.9	78	10	NYR
Nate Schnarr		64	82	19.1	74	10	None
Dmitri Samorukov	64	79	19.1	74	10	MTL
Griffin Mendel		65	83	19.5	75	10	WIN
Josh Anderson		65	69	19.9	75	10	None
Jack McBain		63	79	18.6	73	10	SJ
Kasper Kotkansalo	64	88	19.7	74	10	WIN
Alex Formenton		64	76	18.9	74	10	None
Ruslan Ishkakov		67	78	18	77	10	CAR
Robbie Stucker		67	79	19.9	77	10	MTL
Ivan Muranov		62	78	18.8	72	10	CAR
Adam Thilander		64	76	19.9	74	10	WIN
Kevin Elgestål		52	98	22.2	62	10	None
Samuel Fagemo		59	81	18.4	69	10	NJ
Morgan Barron		64	77	19.7	74	10	CGY
Joshua Burnside		55	90	23.5	65	10	None
Jan Jeník		59	84	17.9	69	10	NYR
Paul Cotter		59	78	18.7	69	10	NJ
Allan McShane		56	79	18.5	66	10	NYR
Liam Kirk		56	75	18.4	66	10	NYR

[VancouverCanucksGM]

Damn, was hoping Abols would boost one more time 8-(