Great analysis except something very very minor, Nam competed at Skate Canada.
Grand Prix Final Analysis
- Thread starter lusterfan
- Start date
Great analysis except something very very minor, Nam competed at Skate Canada.
- Joined
- Mar 3, 2018
- Country
-
They can both make the Final still, if T/M win Rostelecom and MT/M win silver at NHK.I really wanted to see both T/M and MT/M at the GPF, is that still possible?
I ran a little python simulation for ladies. It is debatable, i know 
For 1000 simulated finals, the chance of a skater to be in there is:
Trusova: 100%
Kostornaia: 99.8%
Scherbakova, Kihira: 91.6 и 91.5%
Zagitova: 89.8% (if France was an accident and she will be scored same as last year)
Tuktamysheva: 58.3%
Miyahara: 50.7%
You: 15.6%
Bell, Tennell, Samodurova: 2% or less
Everybody else: 0
If Zagitova will be scored harshly, it becomes this:
Trusova: 100%
Kostornaia: 99.7%
Kihira: 95.4%
Scherbakova: 91.5%
Zagitova: 80.8%
Tuktamysheva: 61.3%
Miyahara: 53.6%
You: 15.1%
Bell, Tennell: 2% or less.
Everybody else: 0
notebooks can be downloaded here: https://gofile.io/?c=pjxOjs (jupyter + python)
My bet for final: Trusova, Kostornaia, Shcherbakova, Kihira, Zagitova, and last one is between Tuktamysheva and Satoko.
For 1000 simulated finals, the chance of a skater to be in there is:
Trusova: 100%
Kostornaia: 99.8%
Scherbakova, Kihira: 91.6 и 91.5%
Zagitova: 89.8% (if France was an accident and she will be scored same as last year)
Tuktamysheva: 58.3%
Miyahara: 50.7%
You: 15.6%
Bell, Tennell, Samodurova: 2% or less
Everybody else: 0
If Zagitova will be scored harshly, it becomes this:
Trusova: 100%
Kostornaia: 99.7%
Kihira: 95.4%
Scherbakova: 91.5%
Zagitova: 80.8%
Tuktamysheva: 61.3%
Miyahara: 53.6%
You: 15.1%
Bell, Tennell: 2% or less.
Everybody else: 0
notebooks can be downloaded here: https://gofile.io/?c=pjxOjs (jupyter + python)
My bet for final: Trusova, Kostornaia, Shcherbakova, Kihira, Zagitova, and last one is between Tuktamysheva and Satoko.
- Joined
- Feb 17, 2010
I ran a little python simulation for ladies. It is debatable, i know
For 1000 simulated finals, the chance of a skater to be in there is:
Trusova: 100%
Kostornaia: 99.8%
Scherbakova, Kihira: 91.6 и 91.5%
Zagitova: 89.8% (if France was an accident and she will be scored same as last year)
Tuktamysheva: 58.3%
Miyahara: 50.7%
You: 15.6%
Bell, Tennell, Samodurova: 2% or less
Everybody else: 0
If Zagitova will be scored harshly, it becomes this:
Trusova: 100%
Kostornaia: 99.7%
Kihira: 95.4%
Scherbakova: 91.5%
Zagitova: 80.8%
Tuktamysheva: 61.3%
Miyahara: 53.6%
You: 15.1%
Bell, Tennell: 2% or less.
Everybody else: 0
notebooks can be downloaded here: https://gofile.io/?c=pjxOjs (jupyter + python)
My bet for final: Trusova, Kostornaia, Shcherbakova, Kihira, Zagitova, and last one is between Tuktamysheva and Satoko.
Out of curiosity, how do 7 skaters have a 50%+ chance of making the final?
- Joined
- Mar 26, 2014
A simple answer is because others have much less than 50%. Let's consider an easy example. Scherbakova, Trusova, Kostornaya, and Sotskova are competing for 2 spots. We assume that the first 3 have equal chances and Maria has 0. Then there are only 3 possible outcomes: S+Tr, S+K, and Tr+K. Each skater is present in 2 of 3. Then 3 skaters have chance of 66.6% to get those 2 spots.
- Joined
- Jun 21, 2003
Another way to think about it is this. One skater making the finals does not prevent another skater from making the finals, too. In Samskrut's example, Trusova making the finals (67% chance) does not reduce Scherbakova's chances to 33%, only to 50%.
Notice that in Samskut's example the probabilities add up to 2. It is Ok for the probabilites add up to more than 100% because we are counting the same outcome twice (S+Tr counts towards S's 67% and also towards Tr's 67%). What can't happen, in Moriel's example, is that the probabilities of 7 skaters adds up to more than 6.)
They do not make it simultaneously.
For example, suppose that we have the following:
Trusova, Kostornaia, Shcherbakova, Kihira make the GPF.
Then, we could have:
1. Zagitova + Tuktamysheva
2. Zagitova + Satoko
3. Satoko + Tuktamyseva
if each of those are equally likely (33% each), then Zagitova, Tuktamysheva and Satoko have 66% chance each.
- Joined
- Jun 21, 2003
Here is another way to look at it. Setting aside the rules (in the simplified example), a priori there are 8 mutually exclusive possibilities:
Z+T+S all three make it.
*Z+T but not S
*T+S but not Z
*S+Z but not T
S but not Z or T
Z but not T or S
T but not S or Z
No one makes it.
Now add in the rules: Exactly two must make it and one not. Under the rules, the probabilities of all combinations except the *-ed ones are 0. The three marked with a * all have probabilty 33%. All 8 probabilities add up to 100%, as expected.
Z+T+S all three make it.
*Z+T but not S
*T+S but not Z
*S+Z but not T
S but not Z or T
Z but not T or S
T but not S or Z
No one makes it.
Now add in the rules: Exactly two must make it and one not. Under the rules, the probabilities of all combinations except the *-ed ones are 0. The three marked with a * all have probabilty 33%. All 8 probabilities add up to 100%, as expected.
Last edited:
I ran a little python simulation for ladies. It is debatable, i know
For 1000 simulated finals, the chance of a skater to be in there is:
Trusova: 100%
Kostornaia: 99.8%
Scherbakova, Kihira: 91.6 и 91.5%
Zagitova: 89.8% (if France was an accident and she will be scored same as last year)
Tuktamysheva: 58.3%
Miyahara: 50.7%
You: 15.6%
Bell, Tennell, Samodurova: 2% or less
Everybody else: 0
If Zagitova will be scored harshly, it becomes this:
Trusova: 100%
Kostornaia: 99.7%
Kihira: 95.4%
Scherbakova: 91.5%
Zagitova: 80.8%
Tuktamysheva: 61.3%
Miyahara: 53.6%
You: 15.1%
Bell, Tennell: 2% or less.
Everybody else: 0
notebooks can be downloaded here: https://gofile.io/?c=pjxOjs (jupyter + python)
My bet for final: Trusova, Kostornaia, Shcherbakova, Kihira, Zagitova, and last one is between Tuktamysheva and Satoko.
Love the math analysis! It's fair for fans to use their heart when making judgments, but for good or for bad, sport is determined by numbers most of the time. It's fun to get nerdy
I don't currently have Python, so I'm curious as to how your simulation was made. If it uses historical scores, I hope you also gave more weight to recent scores, since a pure average would not take into account recency of the scoring environment. Bell is scoring better than she ever has. Satoko has a much better resume and average, but her recent scores (plus the trend of underrotation calls) makes me hesitant to rank her highly. If she can be consistent and not underrotate though, I'd trust her GPF chances. Similarly for Tuktamysheva, she was scored very highly last year in WTT (insane scores there) - she's getting nowhere near that even with an extra 3A. Her 3A is no longer rare this season, and her weird planned jump content is lowering her scoring potential. The simulation might not take that into account. Young You only hit that 3A well once, but she looks sharp and consistent, whereas prior competitions were lackluster.
All I can say is that Cup of China in a few days will be very revealing regarding GPF chances and how the rest of the season will look!
- Joined
- Mar 15, 2018
With the crazy CoC, are Tennell's chance higher?
- Joined
- Mar 15, 2018
It looks like Young and Liza is out of the Final?
It’ll probably be between Bradie and Satoko now...
If Satoko finishes on the podium next week... (highly possible imo) she will be ahead of Bradie and will probably take the last spot as long as NHK will have Alena, Alina and Rika on the podium as expected...
Edit: Sorry I forgot about Mariah... she can make it when finishing second next week with a higher total score than Satoko in case Satoko will be 3rd...
If Satoko finishes on the podium next week... (highly possible imo) she will be ahead of Bradie and will probably take the last spot as long as NHK will have Alena, Alina and Rika on the podium as expected...
Edit: Sorry I forgot about Mariah... she can make it when finishing second next week with a higher total score than Satoko in case Satoko will be 3rd...
- Joined
- Oct 22, 2019
I've been lurking in this thread nervously to see if Bradie is still in it!!! I really really really want Bradie to make it!
I think Bradie's combined total scores are higher than Young, who wins over Liza on placements.
- Joined
- Mar 16, 2017
barring some serious meltdown, for ladies it will be:
1. A. Trusova
2. A. Kostornaia
3. A. Shcherbakova
4. R. Kihira
5. A. Zagitova
6. S. Miyahara
alt
1. B. Tennell
2. E. Tuktamysheva
3. M. Bell/Young You
im assuming that Satoko will get on podium at CoR and Bell is 4th or so. Bradie will be in if Satoko is 4th and score lower than ~217. Anything can happen though, there is also Konstantinova and Yokoi who could play spoiler in CoR
1. A. Trusova
2. A. Kostornaia
3. A. Shcherbakova
4. R. Kihira
5. A. Zagitova
6. S. Miyahara
alt
1. B. Tennell
2. E. Tuktamysheva
3. M. Bell/Young You
im assuming that Satoko will get on podium at CoR and Bell is 4th or so. Bradie will be in if Satoko is 4th and score lower than ~217. Anything can happen though, there is also Konstantinova and Yokoi who could play spoiler in CoR
The Mens CoC results didn't change any of the GPF standings
the results are non relevant. lol
left in the running excluding Hanyu of course
Nam Nguyen
Alexander Samarin
Dmitri Aliev
Kevin Aymoz
Jason Brown
whomever is out of podium between Samarin, ALiev and Nam at CoR will be left out, also Jason Brown has a Japan curse so he probably wont do well in NHK. Kevin Aymoz has the easiest path with only Brown and Hanyu as competitors.
the results are non relevant. lol
left in the running excluding Hanyu of course
Nam Nguyen
Alexander Samarin
Dmitri Aliev
Kevin Aymoz
Jason Brown
whomever is out of podium between Samarin, ALiev and Nam at CoR will be left out, also Jason Brown has a Japan curse so he probably wont do well in NHK. Kevin Aymoz has the easiest path with only Brown and Hanyu as competitors.
