The Rolling Probabilistic Ddp Scoreboard

[gcreptile 10,979]

Update for today's bonanza plus Charles Bradley and other people I am currently forgetting, but they might come back soon. (Edit: Yisrael Kristal)

 

1. David Quantick's Showbiz Pals (190,53) (+1)

2. Thomas Jefferson Survives (190,03) (+1)

3. Golden Slumbers (188,74) (-2)

4. Still Life (183,645) (---)

5. drollercoaster (174,34) (+2)

6. To Kill a Gabor Sister (173,33) (-1)

7. The Love Boat (172,93) (-1)

8. Deceased Hose (167,45) (+1)

9. Poochie died on the way back to his home planet (165,26) (-1)

10. Buckets of Blood (162,82) (---)

11. I will not die, it's the world that will end! (160,32) (+8)

12. Dead Ends (160,145) (---)

13. The B-Team (159,83) (---)

14. Day in the Death (159,8) (---)

15. Heading Nowhere (159,55) (-4)

16. Crossing the Styx (159,47) (-1)

17. I'm sorry for your trouble (158,745) (-1)

18. New Years and Drinks All Night (157,52) (---)

19. The Living End (155,01) (-2)

20. Dead as a Doornail (154,57) (---)

 

Two big winners of this update, in my opinion. TJS with the double slam of Bradley and Booth is now the most serious comeptition of DQSP. The competition has changed from basically Rayya Elias vs. either Cryne or Michael to.... King Michael, yes or no.... He is the most important person for the rest of competition .... unless Greg Gilbert has a ...very... sudden health downturn. 

The other big winner is "I will not die, it's the world that will end!". He (?) also had Booth and Bradley and gains 8 places with them.

 

As we are about the enter the final quarter of the year. The next update might see another reduction in score of those not-terminally ill based picks. Time is running out for them, and without a change in their health status, the probability of them still dying this year is diminishing. But as a contrast to the first reduction in score at half-time, I might also reduce the score of the terminally ill, but not dead people. Maybe by 25%. So those with Leah Bracknell or Gord Downie will suffer. I mean, I already them at very high calculated scores, which I think was in line with everyone's expectations. But now that time is running out, it is time to reflect their stubborn survival. I think it will also help to kick my main team out of the top ten. It will reduce the bias I initially, and almost inevitably had, in favor of my own team. But as my main team drops, so will my B-Team rise (since it doesn't have either Bracknell nor Downie, nor King Michael, nor Rayya Elias, nor Greg Gilbert).

 

[themaninblack 2,112]

According to my calculations, TJS now cannot win the 2017 DDP as each of his remaining picks are covered by the top two in any eventuality. There is no possible combination that enables TJS to pull ahead of the top two...

[Joey Russ 7,229]

8 minutes ago, themaninblack said:

According to my calculations, TJS now cannot win the 2017 DDP as each of his remaining picks are covered by the top two in any eventuality. There is no possible combination that enables TJS to pull ahead of the top two...

There's actually a way for him to pull out on top, but it'll require all 4 of his differentials in some way (Elias, King Michael, Ricketts, and Stefansson) to die. But that is very unlikely...

[themaninblack 2,112]

Just now, Joey Russ said:

There's actually a way for him to pull out on top, but it'll require all 4 of his differentials in some way (Elias, King Michael, Ricketts, and Stefansson) to die. But that is very unlikely...

Lemmme work that one out....

 

Yep, your right. So mathematically, TJS can still win....

[gcreptile 10,979]

27 minutes ago, themaninblack said:

According to my calculations, TJS now cannot win the 2017 DDP as each of his remaining picks are covered by the top two in any eventuality. There is no possible combination that enables TJS to pull ahead of the top two...

Apart from that exception, you're right though. I was basically thinking that it's between TJS and DQSP now, with them having all picks the same except for Gilbert and Michael, so that Michael would give TJS the win. But true, I forgot about GS who would pull ahead with King Michael. So it remains Cryne and/or Michael vs. Elias, though in this case, there is one more differential for Spade: Stefan Karl... in contrast to Greg Gilbert, I could see him still going down this year.

[Death Impends 8,005]

I did a quick check and am 6 points behind both Spade and Bert now so I could win if Rayya and Michael were the only remaining "competitive" hits (+8 for Bert, +9 for Spade, +17 for me). But again, I think Cryne is probably a goner this year, so such a scenario is unlikely. IIRC my ceiling is higher than DQSP and GS, so it's technically never mathematically impossible for me (but I think Karl and Ricketts see 2018).

[gcreptile 10,979]

6 minutes ago, Death Impends said:

I did a quick check and am 6 points behind both Spade and Bert now so I could win if Rayya and Michael were the only remaining "competitive" hits. But again, I think Cryne is probably a goner this year, so such a scenario is unlikely. IIRC my ceiling is higher than DSQP and GS, so it's technically never mathematically impossible for me (but I think Karl and Ricketts see 2018).

Yes, your ceiling is 213, DQSP's is 212. GS's is 205.

Also, in my opinion, your most likely survivor (Ricketts) is "only" worth 9 points, and Spade's most likely survivor (Gilbert) is 10pts. So you'd also win the hypothetical 19/20 scenario.

 

Personal dream scenario:

Herbert Kretzmer dies in 2 weeks (his birthday) giving The B-Team a temporary lead of 2 pts....

[Joey Russ 7,229]

10 minutes ago, gcreptile said:

Also, in my opinion, your most likely survivor (Ricketts) is "only" worth 9 points, and Spade's most likely survivor (Gilbert) is 10pts. So you'd also win the hypothetical 19/20 scenario.

Ricketts is worth ten points since he was born in 1968, no?

[msc 18,480]

Ah but if Pan Breed sweeps their entire board (everyone dies, obits for the 4 dead non-obit folk show up) then it's 220 points. Be afraid, everyone...

 

 

[gcreptile 10,979]

13 minutes ago, Joey Russ said:

Ricketts is worth ten points since he was born in 1968, no?

Hm, indeed. Could be I initially calculated that AFTER I had already switched my longlist from "2016" to "2017", so that Ricketts, picked in 2017 for 2018, is possibly only worth 9 pts. Then the ceiling of TJS is 1 point higher. He still wins that 19/20 scenario though.

[Joey Russ 7,229]

12 minutes ago, gcreptile said:

Hm, indeed. Could be I initially calculated that AFTER I had already switched my longlist from "2016" to "2017", so that Ricketts, picked in 2017 for 2018, is possibly only worth 9 pts. Then the ceiling of TJS is 1 point higher. He still wins that 19/20 scenario though.

He could win the 16/20 scenario if Nevin, Bracknell, and Downie sees the year out...

[drol 11,979]

19 hours ago, gcreptile said:

Yes, your ceiling is 213, DQSP's is 212. GS's is 205.

Also, in my opinion, your most likely survivor (Ricketts) is "only" worth 9 points, and Spade's most likely survivor (Gilbert) is 10pts. So you'd also win the hypothetical 19/20 scenario.

 

Personal dream scenario:

Herbert Kretzmer dies in 2 weeks (his birthday) giving The B-Team a temporary lead of 2 pts....

Why do you think Herbert Kretzmer will die this year? Is he ill?

[gcreptile 10,979]

1 hour ago, drol said:

Why do you think Herbert Kretzmer will die this year? Is he ill?

He had aged considerably in a picture in late 2015 and had no public appearances in 2016.

And well, I had a couple of free slots in my B-Team and wanted to fill it with my own "unusual" findings - see also John Swinton of Kimmerghame.

Initially, I had the B-Team full of likely unique picks. I just wanted to get "interesting" hits. Of course, now I know I should have just put John Wetton in there and the team would have been a title contender.

[gcreptile 10,979]

Update for Gord Downie, Daisy Berkowitz, a belated Terri Roberts non-obit and score reductions for every person still alive - because time is running out! I simply reduced the score of every living person by a third. So now those who already have more hits will do better in this than those who still have more upward potential, but fewer actual hits. Those with the joker Leah Bracknell -bad luck! But the eventual bonus, should she still die this year, will be bigger.

 

1. David Quantick's Showbiz Pals (185,93333) (---)

2. Thomas Jefferson Survives (181,67333) (---)

3. Golden Slumbers (180,68) (---)

4. Still Life (178,0333) (---)

5. drollercoaster (163,62) (---)

6. To Kill a Gabor Sister (158,206667) (---)

7. Deceased Hose (154,82) (+1)

8. The Love Boat (154,74) (-1)

9. Buckets of Blood (154,69) (---)

10. Dead Ends (154,08333) (+1)

11. The B-Team (152,936667) (+2)

12. Crossing The Styx (152,90) (+4)

13. Heading Nowhere (151,02) (+2)

14. Day in the Death (150,27) (---)

15. Poochie Died in the way back to his home planet (150,16) (-6)

16. I will not die, it's the world that will end! (149,33333) (-5)

17. Guild of Master Reapers (145,13333) (re-entry)

18. I'm sorry for your trouble (144,816667) (-1)

19. New Years and Drinks All Night (142) (---)

20. Drunkasaskunk (141,53) (---)

 

In spite of the rather dramatic change in the scoring system, not many changes in places (but DQSP is pulling half a pick ahead)! So the method sort of works, but it also shows where I failed the most: My own teams. I undervalued The B-Team and overvalued The Love Boat. But I said since the beginning that I somewhat expected this bias, because if it had been any different, my teams would have been different.

 

Otherwise, I got a couple of things quite right, like last year's no2 and 3 underperforming.

I undervalued Charles Bradley and Paul Van Zandvliet. The former was just inside knowledge that led me astray, the latter was misplaced doubt about his obitability. On the other hand, I overvalued the obitability of Ben Suisala, and possibly of Abu Bakr al-Baghdadi.

 

There will be one more update next month, with another reduction of score of the then still living people by another third.

 

[msc 18,480]

You trying this again next year?

[gcreptile 10,979]

7 minutes ago, msc said:

You trying this again next year?

Maybe. The Excel table that leads to these scores IS my longlist. So I can definitely calculate some quick scores for the regular teams. And will probably do so just for myself anyway. But going through all the several hundred teams to find the favourites eventually making the TOP 20, doing adequate research on the people my longlist missed, and maintaining the live scoreboard is a bit of work. I might try to do it bit by bit but wouldn't commit to it. At the moment, I could rather imagine running one of the regular deadpools if anyone feels like taking a break or something. I'd also be up for some DDP biographical duties. This would take the time I'd otherwise use for my ranking.

[gcreptile 10,979]

Ahem, so that last update will not be coming... I have to apologize, December was terribly busy and it might impact my team selections for 2018. - But then again, maybe not, since I already wanted to avoid hunting for those December names with terminal cancer, like Ricketts and Gilbert. So my research was quite good in mid-November. Let's see how far it takes me....

 

That William H Gass hit would have pushed Bucket of Blood to the place it belongs. The Love Boat would have gotten two more hits with Sharon Laws and Dmitri Hvorostovsky. In some way, it was justified that The Love Boat still scored so high in spite of being out of the Top 20 for most of the year. One more hit and it would have been Top 10. But that hit didn't come. Rayya Elias, Catherine Nevin and especially Leah Bracknell (the joker) lasted an unusual amount of time or exaggerated their illness. 

 

The biggest final score change would have come from King Michael who might have pushed Golden Slumbers to the top. I would also have reduced the possible scores of all still living people by another half.

 

As a general resume: once unique hits and Drop 40 bonusses became clear, I had the top scores in the region of 140 pts. That was obviously too low. I wonder if I can get more precise next year, and will spend a week in January to do that. I know there was systematic bias in my calculations in favor of The Love Boat and against The B-Team. I also estimated the score of msc and Joey Russ - because I gave Ben Suisala a chance of obiting. I'm not sure if I made a real miscalculation there. Or actually, Pan Breed just ended up on the bad side of all my hedges - as you might recall, I also picked Alan Aspin and Ben Suisala in Shameless. So I gave them a much higher score than they received (zero). Then there was Abu Bakr al-Bahgdadi: For 2018, I have significantly reduced the expected score of him and all terrorists. I valued them too highly.

 

On the other hand, I undervalued Golden Slumbers. After Drop 40 and unique bonusses he was 7th. I know, I undervalued Paul Van Zandvliet and completely ignored John Wetton. The difference between him and DQSP was 5 pts, so less than one hit. He won by one hit, and I think it's hard to predict the competition with such precision. All of Spade's survivors are also on the Love Boat, so my bias in favor of my own picks is also slightly present in that team. Golden Slumbers also had Patrick Cryne and Jerome Golmard who I didn't know about before. I had to make up a score for both. Though actually, since one of them died, and one didn't, it probably didn't distort all that much.

There was Charles Bradley who, as you might remember, I knew about but didn't pick him because a german concert organiser told me that he was doing well in his cancer battle. He returned on tour, only to see the cancer return and kill him within weeks. I guess that's one of the things that just happen...

I can pretty much see where about 10-15 missing points in my calculation came from. (Jagraj Singh was another miss. I picked him in the Deathrace, but didn't expect him to obit. He just barely did, with the least credible obit in DDP history....)

I also wonder why Spade didn't pick Paul Van Zandvliet.

 

Ok, enough navel gazing. I will definitely put up the team that my calculation thinks would have been the best. It will most likely not be the one I pick, because I don't want to pick more than 3 cancer mums/dads.

[msc 18,480]

Brackers, Elias and Nevin, eh? In order: still within the average life expectancy (more of us should have twigged that last Christmas, tbh); literally near death all year but hanging on as a massive outlier; exaggerated to all hell.

 

Pan Breed made three terminal errors last year, from what I can see. The first was subbing out Hewlett/Snuka/Magill/Skellern right before I sent my team for Aspin/Baghdadi/Holmlund/Suisala. A point deficit of 41 points there. Very daft. Second, ignoring deadpool sabermetrics (as Spade puts it) on a few picks that weren't going to obit. Third, and most damning, a complete lack of judgement. About 34 names from my final shortlist of 50 are gone, but only 10 from my main DDP team. Says it all.

 

I'm not sure how your system can account for the third one overall, but I will be interested to see!

[Joey Russ 7,229]

When do you think you’ll put the first prediction for 2018? 

[gcreptile 10,979]

On 2.1.2018 at 23:03, Joey Russ said:

When do you think you’ll put the first prediction for 2018? 

Not before next weekend.

But one thing I could quickly do...the best possible team one could have picked according to my system:

 

Marieke Vervoort (joker)

Leah Bracknell

Liam Miller

John McCain

Dean Francis

Rayya Elias

Patrick Cryne

Greg Gilbert

Vanessa Goodwin

Jason Bragg

Bobby Zarin

Devin Lima

Tiffany Youngs

Deborah James

Nobby Stiles

Fernando Ricksen

Peter Lorimer

Tessa Jowell

Stefan Karl Stefansson

Shalom Ouanounou  (Carl Piddington)     That would have been 167,75 "expected" points. (Cas Willow deleted)

[gcreptile 10,979]

Let me see if that works:

	The Love Boat				The B-Team				Going Down in Down Under
	Name	Expected score	base points		Name	Expected score	base points		Name	Expected score	base points
1	Marieke Vervoort	12,49	13,00	1	Morgan Tsvangirai	6,02	8,00	1	John "Cocksy" Cocks	4,54	9,00
1	Marieke Vervoort	12,49	13,00	1	Morgan Tsvangirai	6,02	8,00	1	John "Cocksy" Cocks	4,536	9,00
2	Shalom Ouanounou	6,34	12,00	2	Vanessa Goodwin	7,42	9,00	2	Ted Mack	3,78	6,00
3	Leah Bracknell	10,80	12,00	3	Jean-Louis Trintignant	3,73	6,00	3	Vanessa Goodwin	7,42	9,00
4	Dean Francis	8,05	10,00	4	Philomena Lynott	4,54	5,00	4	Jeanne Little	4,60	7,00
5	Emma Hannigan	5,31	10,00	5	Dawn Wilson	5,81	10,00	5	Sara Chivers	5,60	10,00
6	Thuy Thanh Truong	5,38	10,00	6	Kathleen Blanco	4,20	7,00	6	Kevin "Chalky" Carr	4,58	10,00
7	Sara Chivers	5,60	10,00	7	Nicholas Edwards	4,20	6,00	7	Clive James	6,00	10,00
8	Devin Lima	7,00	10,00	8	Vanessa LaFaye	5,21	7,00	8	Grant Goldman	4,21	11,00
9	Thomas S. Monson	4,32	5,00	9	Dilip Kumar	4,24	5,00	9	Olivia Newton-John	5,50	11,00
10	Rayya Elias	7,53	9,00	10	Genesis P-Orridge	4,89	8,00	10	David Cervinski	3,71	10,00
11	Nobby Stiles	6,79	10,00	11	Denis Goldberg	4,80	6,00	11	Barry Du Bois	5,58	9,00
12	Pauline Maroney	4,59	10,00	12	Peter Lorimer	6,57	7,00	12	Peter Thomson	1,80	6,00
13	Sumner Redstone	4,48	5,00	13	Patrick Cryne	7,47	7,00	13	Stuart Fraser	1,73	8,00
14	Liam Miller	10,80	13,00	14	Satoru Anzaki	3,84	6,00	14	Mike Willesee	2,38	7,00
15	Matt Capotelli	4,64	10,00	15	Franco Zeffirelli	5,60	5,00	15	Conway Savage	2,63	9,00
16	Tiffany Youngs	6,90	10,00	16	Bobby Zarin	7,00	7,00	16	Blazo Vuksanovic	5,39	8,00
17	Fernando Ricksen	6,75	13,00	17	Charlotte Rae	5,85	5,00	17	Donna Penny	5,21	10,00
18	Hector Timerman	6,16	11,00	18	Johnny Hubbard	5,40	6,00	18	Tommy Raudonikis	5,23	8,00
19	John McCain	9,00	9,00	19	Anna Pavord	4,25	7,00	19	Thomas Keneally	2,52	6,00
20	Tessa Jowell	6,37	10,00	20	Lindy Remigino	3,60	6,00	20	Jessica Falkholt	6,15	10,00
All		151,784	215	All		110,648	141	All		93,083	183

 

This is what I have with my teams. This is BEFORE Drop 40 and unique bonusses. Obviously the expected score will change once these bonusses have a probability of either 1 or 0 (i.e. they are known)

 

In the column "base points" you can see my assumptions: For example Marieke Vervoort dies an unnatural death but is not a Drop 40 pick (may be wrong?)

Neither do I have Shalom Ouanounou and Emma Hannigan as unique, but Hector Timerman and Pauline Maroney are.

I also think that Leah, Nobby, Fernando, John and Tessa will be on the Drop 40.

 

Now... these scores also build on the assumption that my sub Thomas S. Monson replaces Derek Dudley (should be no problem), and that Sumner Redstone replaces Cas Willow (Spade????). It could be that I, and anyone else who picked her, lose these points.

 

Edit: I also want to add that some last minute information isn't in these numbers. For example, Jessica Falkholt being described as a 50:50 case, or Hector Timerman writing a New York Times piece in late December (increasing his obitability). I have to take another look at that.

 

Second edit: If you wonder, why is, for example, Charlotte Rae's expected score higher than her base points? Well, it's because I gave her a chance of getting into the Drop 40, but I wasn't sure enough of it to give her 8 base points.

[Death Impends 8,005]

I think Vervoort makes Drop 40. Used Greg Gilbert 2017 as a point of reference - he first landed on the radars about a week and a half before the start line, with across the board UK coverage of his illness. He got very close to the Drop 40 despite being a 50/50 pick. So I think someone (not counting cancer mums, deathbed marriages, coma people etc) who receives the same level of coverage so late in the year (like with Vervoort's Telegraph interview) while being a dead cert, makes the cut.

[viGnoS 18]

I don't understand how you are calculating the "expected score" ? is that based on a "dying probability" ??

some expected score are over base point so their dying probability are over 100% ? sounds crazy !

[gcreptile 10,979]

9 minutes ago, Death Impends said:

I think Vervoort makes Drop 40. Used Greg Gilbert 2017 as a point of reference - he first landed on the radars about a week and a half before the start line, with across the board UK coverage of his illness. He got very close to the Drop 40 despite being a 50/50 pick. So I think someone (not counting cancer mums, deathbed marriages, coma people etc) who receives the same level of coverage so late in the year (like with Vervoort's Telegraph interview) while being a dead cert, makes the cut.

I think you will be right. I had a strange fixation on trying to surpass my base points of 2017. (216? 220? something like that) The base points are just another variable. In Vervoort's expected score (12,49) is a 20% chance of making the Drop 40. But I can already see that this was too low, probably more like 60-70%. Those who have her as joker may well have a one-pick advantage over those who haven't.

[msc 18,480]

If she goes she might be a 32 point joker.