(20 January 2022) 

In a previous discussion we compared the performance history of the top men who will be competing at the Olympic Winter Games in Beijing.  Using the top level scores the numbers suggest the Chen / Hanyu matchup is pretty much neck and neck.  Based on their scores over the past year, if both skaters turn in their best Chen wins.  If both skaters turn in their average Chen wins.  If both skaters skate their worst Hanyu wins.  Or in other words, Hanyu's chances are tied to whether Chen makes significant errors or not.  Drilling down into their program content it's clear why.

Short Programs

In the short programs it comes down to the three jump elements:  4F, 3A and 4Lz+3T (in second half) for Chen, and 4S, 4T+3T and 3A (in second half) for Hanyu.  Thus, Hanyu starts out with a 4.07 point deficit in base value in the short program that he has to make up in GoEs and program components.  This is not an easy task.  It means his short program has to be near perfect against a less than perfect Chen.  For example, to make up a 4 point deficit Hanyu would have to outscore Chen by 1 GoE for each element, or outscore Chen by 0.8, in each component, or some mix thereof.  Past results say the best Hanyu might hope to outscore Chen in program components is 0.4-0.6 in each, or 2-3 points total, on a bad day for Chen.

Free Skates

An even bigger obstacle faces Hanyu in the free skate when you compare base values from the Japanese and U.S. national championships.  In those competitions Chen had a base value of 101.24 to 88.40 for Hanyu, for a deficit of 12.84 for Hanyu.

This deficit comes primarily from three elements: Hanyu's downgraded 4A attempt, 3A+2T instead of 3A+3T and 3Lo instead of 4Lo.  Further, Hanyu's base value deficit all year has exceeded 10 points: 14.14 at 2021 Worlds and 10.10 at 2021 World Team Trophy compared to Chen at the same competitions.  Also in the base value deficit (at 1.1 point in the second half) is Hanyu's 4T+1Eu+3S against 4T+1Eu+3F for Chen.

Making up that many points in GoEs and program components is nearly impossible.  It would require scoring 1½ GoE higher in every element, or scoring up to 1.4 points higher in every program component mark - an impossible task.  Even a mix of those two will be difficult.  Given that Hanyu already scores 3-4s in GoEs (or better), increasing by 1½ GoEs essentially means scoring perfect 5s for all elements, and as we see below it still would not be enough.

The quad Axel attempt adds to Hanyu's problem.  Although 4A<< has a respectable base value of 8.0, if it is scored at -5 he loses an additional 4 points, and if he falls he loses another point.  The best he can hope for by the rules of scoring is to lose 0.8 points on the attempt if the jump is spectacular and is only downgraded, and not also two footed as it was at Japanese Nationals.

Since Hanyu does not have a quad flip or Lutz to punch up his base value, he would need to convert 3A+2T to 3A+3T and 3Lo to 4Lo.  But since he is already repeating quad toe loop and triple Axel, and has a triple toe loop in combination with quad toe loop, he can't put a triple toe loop on one of the triple Axels.

Turning the triple loop into a quad buys him 5.6 base value points of the 10-14 he will likely need.  That would then free up the triple loop to add 3A+3Lo instead of 3A+2T, but as far as we know that combination is not in his arsenal.  If it were, that would buy him another 3.6 points in base value.  Together these hypothetical changes would get him within striking distance of a win with superior GoEs and program components, such as half his elements scored 1 GoE higher than Chen and his program components scored 0.25 points higher.

Hanyu is basically stuck between a rock and a hard place for jump content because he does not use quad flip or quad Lutz and the quad loop is insufficiently reliable that he often does not use it even once in a program no less twice.  A reliable quad loop would allow him to repeat the quad loop instead of the triple Axel to close the gap in base value by 2.5 points.

To further compare Hanyu's scoring possibilities we look at some synthetic protocols which we compare to likely Chen scores.

If Chen is at the top of his game in Beijing, we expect he will perform at least as well as he did at U.S. Nationals, with the addition of not falling on quad flip and the choreographic sequence.  We also adjust his GoE points downward by 5% to account for the usual marks inflation at national championships. The result is a best case total element score of 134 that Hanyu will have to overcome, compared to the 125.89 TES that Chen scored at 2021 Worlds.

Table 1.  Likely best case for Chen technical scores.

In terms of the TES Hanyu needs to achieve, one should note that if GoE inflation at U.S. Nationals is 10% rather than the assumed 5% the target for Hanyu is more like 132 points.  For the following, nevertheless, we will assume the 133.62 points as the TES target for Hanyu.

Next we show how close Hanyu came to Chen's scores at 2021 Japanese Nationals.

Table 2.  Technical scores for Hanyu at 2021 Japanese Nationals.

Hanyu comes up nearly 20 points short in TES.  The lesser base values hurt the skater two ways, in the base value itself, and in the GoE points being a percentage of the base value.  For the same GoEs, the element with the higher base value also gains the higher GoE points.

Even with all 5s for GoEs the content in Table 2 is not competitive with Chen, as seen next.

Table 3. Technical scores for Hanyu with Japanese Nationals content and +5 GoEs.

Element	Base		GoE Points	Ave GoE	Points
4A<<	8.00		-3.89	-4.86	4.11
4S	9.70		4.85	5.00	14.55
3A+2T	9.30		4.00	5.00	13.30
3Lo	4.90		2.45	5.00	7.35
FCCoSp4	3.50		1.75	5.00	5.25
StSq4	3.90		1.95	5.00	5.85
4T+3T	15.07	X	4.75	5.00	19.82
4T+1Eu+3S	15.73	X	4.75	5.00	20.48
3A	8.80	X	4.00	5.00	12.80
ChSq1	3.00		2.50	5.00	5.50
FCSSp4	3.00		1.50	5.00	4.50
CCoSp4	3.50		1.75	5.00	5.25
	88.40		30.36	TES	118.76

This content is a losing hand even with perfect program component scores.

Next we present several scenarios for changes to the element content to increase the base value of the program for Hanyu. Some of these changes are within his arsenal and some are not.

Table 4.  Hanyu hypothetical scores with one quad loop.

Element	Base		GoE Points	Ave GoE	Points
4A<<	8.00		-4.00	-5.00	4.00
4S	9.70		4.37	4.50	14.07
3A+2T	9.30		3.60	4.50	12.90
4Lo	10.50		4.73	4.50	15.23
FCCoSp4	3.50		1.58	4.50	5.08
StSq4	3.90		1.95	5.00	5.85
4T+3T	15.07	X	4.28	4.50	19.35
4T+1Eu+3S	15.73	X	4.28	4.50	20.01
3A	8.80	X	3.60	4.50	12.40
ChSq1	3.00		2.50	5.00	5.50
FCSSp4	3.00		1.35	4.50	4.35
CCoSp4	3.50		1.58	4.50	5.08
	94.00		29.79	TES	123.79

In this scenario Hanyu is nearly 10 points short in TES, even with 4s and 5s for all GoEs, a deficit that cannot be made up in program component scores for a clean Chen skate.

Table 5.  Hanyu hypothetical scores with one quad loop and triple Axel + triple loop.

Element	Base		GoE Points	Ave GoE	Points
4A<<	8.00		-4.00	-5.00	4.00
4S	9.70		4.37	4.50	14.07
3A+3Lo	12.90		3.60	4.50	16.50
4Lo	10.50		4.73	4.50	15.23
FCCoSp4	3.50		1.58	4.50	5.08
StSq4	3.90		1.95	5.00	5.85
4T+3T	15.07	X	4.28	4.50	19.35
4T+1Eu+3S	15.73	X	4.28	4.50	20.01
3A	8.80	X	3.60	4.50	12.40
ChSq1	3.00		2.50	5.00	5.50
FCSSp4	3.00		1.35	4.50	4.35
CCoSp4	3.50		1.58	4.50	5.08
	97.60		29.79	TES	127.39

In this scenario Hanyu is nearly 6 points short in TES, even with 4s and 5s for all GoEs, except for the quad Axel.  It requires program component scores significantly higher than Chen.  As far as we know Hanyu does not actually have a 3A+3Lo.

Table 6. Hanyu hypothetical scores with two quad loops.

Element	Base		GoE Points	Ave GoE	Points
4A<<	8.00		-4.00	-5.00	4.00
4S	9.70		4.37	4.50	14.07
4Lo+2T	11.80		4.73	4.50	16.53
4Lo	10.50		4.73	4.50	15.23
FCCoSp4	3.50		1.58	4.50	5.08
StSq4	3.90		1.95	5.00	5.85
4T+3T	15.07	X	4.28	4.50	19.35
4T+1Eu+3S	15.73	X	4.28	4.50	20.01
3A	8.80	X	3.60	4.50	12.40
ChSq1	3.00		2.50	5.00	5.50
FCSSp4	3.00		1.35	4.50	4.35
CCoSp4	3.50		1.58	4.50	5.08
	96.50		30.92	TES	127.42

In this scenario Hanyu is nearly 6 points short in TES, even with 4s and 5s for all GoE, except for the quad Axel.  It requires program component scores significantly higher than Chen.  This scenario requires six quads to Chen's five, a stressing and unlikely scenario.  An alternate to this scenario is to use 4S+2T as the third element for a reduced TES of 126.26.

In the above scenarios we put the value of the quad Axel at 4 points, which assumes a downgraded and two-footed jump as executed at Japanese Nationals.

Using the known jumps in Hanyu's arsenal, the following content puts him even with Chen in TES, if rewarded with outstanding GoEs.

Table 7.  Hanyu hypothetical scores using jumps no more difficult than quad loop.

3Lz	5.90		3.60	4.50	9.50
4S	9.70		4.37	4.50	14.07
4Lo+2T	11.80		4.73	4.50	16.53
4Lo	10.50		4.73	4.50	15.23
FCCoSp4	3.50		1.58	4.50	5.08
StSq4	3.90		1.95	5.00	5.85
4S+3T	15.40	X	4.37	4.50	19.77
4T+1Eu+3F	16.83	X	4.28	4.50	21.11
3A	8.80	X	3.60	4.50	12.40
ChSq1	3.00		2.50	5.00	5.50
FCSSp4	3.00		1.35	4.50	4.35
CCoSp4	3.50		1.58	4.50	5.08
	95.83		38.61	TES	134.44

This content trails Chen 5.41 points in base value, an amount that can be made up with a realistic combination of higher GoEs and program component scores.

Quad Axel to the Rescue?

If Hanyu's quad Axel improves by the Olympics, to say 4A<, the value of that element would go from 4 points to 5 to 13 points, depending on the GoE (-5 to +3 maximum for an under-rotation with no other errors), so a potential gain in the above scenarios of 1 to 9 points.  The example in Table 4 with 4A< is a potential route to victory for Hanyu.  So also is 4A<< that is otherwise spectacular and not two-footed, which could be worth up to 7.2 points instead of the current 4.  Both of these possibilities still require Hanyu outscoring Chen in program components by several, but achievable, points.

Finally, if the quad Axel is abandoned and replaced with the highest base value triple available, with reasonable execution (3Lz at +3 GoE), a 4 point element becomes a 7.67 point element, a gain of 3.67 points.  This also makes Table 4 with this substitution a potential route to victory and with less risk.

Table 8.  Point comparison for Hanyu element 1.

One final factor weighing on Hanyu's chances is that he must avoid serious errors. With one serious error program components cannot be 10s, but still can be high enough to outscore Chen in components. With two or more serious errors, program component scores are capped at values that making winning impossible without Chen totally imploding.

Summary

If Chen skates his best in Beijing he can be expected to score between 126 and 134 Total Element Points (or perhaps 132 depending on the degree of GoE inflation at U.S. Nationals).

For Hanyu to win by not relying on Chen making major errors he must:

1. be within striking distance of a win after the short program (1-2 points).  This will most likely require outscoring Chen in program components in the short by several points.
2. add at a minimum a well executed quad loop to the free skate.
3. have superior GoEs for all elements, with an even mix of all 4s and 5s or better.
4. execute a spectacular 4A<< attempt with no other errors OR increase the rotation of the quad Axel to at least 4A<, with no other errors, OR replace it with a well executed triple Lutz (or even triple flip).
5. outscore Chen in program components by a significant, though obtainable, amount in both programs.

Hanyu does not need a fully rotated quad Axel to bring his total element score within striking distance of Chen in program components.  It only has to be sufficiently less crappy than a completely failed execution to earn points equivalent to a lesser value quad.

There is, of course, no guarantee Chen will skate error free in Beijing.  As Carlo Fassi used to say, and others since over the years, 'the ice is very slippery,'  but counting on your competition to make errors in order to win is a desperate strategy for victory, or no strategy at all.

Chen's higher base value content makes him the unambiguous favorite for the gold medal.  Hanyu's chances for victory without Chen making a serious error in his programs is exceedingly small, though not zero.  And perhaps he will fool us all and significantly rework his content from Japanese Nationals to Beijing.  In any event, the Men's competition in Beijing is sure to be the most dramatic and compelling of the Games.