Should Falls be Penalized More Severely?

[Tak]

I dont intend this thread to be a re-hash of complaints about specific skaters, but rather to look at what could or should be done to change the judging parameters [if anything]. [So please let's not talk about specific skaters, but rather the judging system in general].

Another poster on another thread said the problem was falls need to be more severely penalized. I agree with this. Another poster said multiple falls should increase the penalty on each successive fall. This is an interesting idea.

My own suggestion is to keep the -1 penalty for each fall, but to include an automatic downgrade on the jump. What I mean is this - if a skater falls on a quad, and the caller calls it as a quad, it is downgraded to a triple because of the fall. If the caller calls it as a triple, it gets downgraded to a double, etc.

Does anyone have any thoughts?

[Marco]

1. Someone scores high with fall on difficult jumps

2. People complain

3. Falls are penalized more

4. Less people try hard jumps

5. Competitions are won by conservative skates

6. People complain

7. Jump values are raised again

8. go back to 1

The circle of life

[Vash01]

The problem with the current penalties on falls is that it is an extremely small percentage of the total score. Typical scores at higher levels are in the 80's or 90's for SP, and in the high 100's for the LP. So I am going to approximate them at 100. When a skater loses a mere 1 point as a result of the fall (and receives negative GOEs) it is still a very small fraction (1% or less) of the total score. In a very close competition this CAN make a difference, but when that is not the case, the difference in the technical marks is not that big. It allows the judges to manipulate the PCS marks to favor certain skaters with higher reputation. So I would like to see bigger penalties for falls, and either increasing penalties for successive falls, or a way to change (or better defined) the PCS system when there are multiple falls in a performance.

I am with the poster who wrote that the technical marks should have more weight than the PCS marks, because it is easier to manipulate the PCS scores. There still will be the issue of the technical expert calling the elements correctly, so there is no perfect system in a subjective sport like FS, but what we have now needs more improvement. JMO.

[AndyWarhol]

Yes, there should be a bigger deduction for falls in my opinion. It makes it more exciting as well, as there is real risk involved in going for those jumps, and if it means the men are only trying one quad per program then be it. Its not much use upping the technical if the world champion can't even land their planned content.

[fscric]

Haha! It's getting more and more comical.

[VIETgrlTerifa]

Weren't some of the bigger issues for the more conservative skates in the 2007-2010 quad were that people didn't think that the quad was valued high enough, UR penalties more severe (there wasn't a differentiation between > and >>), and that GOEs had a much higher value for jumps? All those factors meant that skaters could get away without taking as much of a risk as long as they completed fully-rotated jumps? It seemed that a quad wasn't worth it because a devalued quad or putting all that energy into the quad didn't seem to garner a big enough reward.

Maybe, if the code continued to value the quads on that high of a level, but had a harsher penalty for falls so that it could "more accurately" calculate how much the disruptions affected the performances, then skaters would still want to go for the risk of a quad. Maybe one fall should be left alone, but once it gets into multiple falls or stumbles, then it should have some sort of effect on the overall score.

Yes, I know that's an overly general proposition and it does not truly balance out the risk v. conservative problem, but maybe it's a start.

[kosjenka]

As far as deductions go - no.
P/E should reflect the actual performance, not what skaters show in practice or during other competitions.

[eleonorad]

1. Someone scores high with fall on difficult jumps

2. People complain

3. Falls are penalized more

4. Less people try hard jumps

5. Competitions are won by conservative skates

6. People complain

7. Jump values are raised again

8. go back to 1

The circle of life

True. But I prefer stopping at n.7. There's nothing wrong with the current system. The only thing that is wrong here is the way the judges use GOEs and PCS to do politics.

[bek]

1. Someone scores high with fall on difficult jumps

2. People complain

3. Falls are penalized more

4. Less people try hard jumps

5. Competitions are won by conservative skates

6. People complain

7. Jump values are raised again

8. go back to 1

The circle of life

In 6.0 we had people doing quads rewarded but we actually normally didn't have 2 falls getting you a world title. The idea Marco that its somehow impossible to have both is ridiculous.

I think what most people want is multiple mistakes hit more. I would be fine if in P/E the judges were allowed in general to take into account jump content.


Also nobody should be getting 6/7 points for a quad fall. 3 points would be much more fair. (2 points for a fall) Your taking a risk. And for those who say then no quads quads are now worth more in general and they could add in other incentives to the quad like rewarding combos, saying a guy could repeat one quad type and two other triples. All that would reward guys with quads.

All of this can be done without handing someone huge points for quad falls.

[misskarne]

In 2002 SLC you had to be doing three quads at LEAST (one in the short, two in the long) to be competitive, but two falls would have killed your competition.

[BittyBug]

Yes, falls should be penalized more harshly. You fall on an element, you should get no credit for that element - it really doesn't have to be more complicated than that.

[TheIronLady]

I think falls in skating, as Dick Button used to tell us, are more acceptable when they do not disrupt the flow of the program. I want to point out the idea was around for a while that you could fall on significantly difficult elements and still complete a wonderful program.

[morqet]

There was a lot of discussion about this on twitter after the Men's FS - I think a lot of people ended up agreeing around the idea that if you fall on an element the base value is reduced by a certain percentage (but no agreement on what that percentage is! I'd go for 50% myself), then you apply goe as usual. If you fall on a jump you have failed to complete the element properly, & it seems ridiculous that skaters can still get 6/7 points for a "failed" element.

There also needs to be more consideration of how falls affect PCS scores, but I don't think you can build mandatory deductions into P/E. For me, some falls like MBM's in her free program or Kostner's in her short don't disrupt the whole performance, but some, like Chan's in his free, really detracted from the program. Judges should adjust the PCS marks they give accordingly - and this is something that isn't done at the moment - but I think there are too many variables to be able to have mandatory PCS deductions.

There's also a pretty interesting analysis of the scoring system including penalties for falls here.

[Mafke]

My proposal:

Fall on a scoring element : no penalty but no points earned for that element (obviously some fine tuning would be needed for falls on the second jump in a sequence/combination)

Fall on a non-scoring element : -3 (or more) points

[bek]

There was a lot of discussion about this on twitter after the Men's FS - I think a lot of people ended up agreeing around the idea that if you fall on an element the base value is reduced by a certain percentage (but no agreement on what that percentage is! I'd go for 50% myself), then you apply goe as usual. If you fall on a jump you have failed to complete the element properly, & it seems ridiculous that skaters can still get 6/7 points for a "failed" element.

There also needs to be more consideration of how falls affect PCS scores, but I don't think you can build mandatory deductions into P/E. For me, some falls like MBM's in her free program or Kostner's in her short don't disrupt the whole performance, but some, like Chan's in his free, really detracted from the program. Judges should adjust the PCS marks they give accordingly - and this is something that isn't done at the moment - but I think there are too many variables to be able to have mandatory PCS deductions.

There's also a pretty interesting analysis of the scoring system including penalties for falls here.

The problem with saying some falls distract and some falls don't. I mean there is a truth to this but there's also a whole lot of subjectiveness to it.

Also I think once you start gettting into multiple falls, than it doesn't matter how many smiles etc the skater is giving, I think the overall impression of the performance IS affected. No mandatory deductions leaves far too many "outs" for the judges to argue the perenial favorite didn't.

Personally I think there needs to be an increased penalty per mistake I.e a great performance can get away with one fall, but I don't think it can get away with multiple.

Now yes I can see someone getting even more hit...

I do think falls in the technical program should count more though... it is the technical program. Also it shouldn't be just falls it should be messiness. Chan had two falls but he also had a really bad stumble on a third jump. (It wasn't like it was a tiny two foot)

[mag]

If we are talking about program disruption, IMHO the celebrating down by some skaters after each jump is way more disruptive to the flow of a program than a fall. And for those who say there were more quads and fewer falls in Salt Lake, well, if the program is stroke, glide on two feet, jump, wash, rinse, repeat, of course you get fewer falls. Spins and much of the footwork done in Salt Lake doesn't cut it in Novice these days.

[gkelly]

Let's think this through.

My own suggestion is to keep the -1 penalty for each fall, but to include an automatic downgrade on the jump. What I mean is this - if a skater falls on a quad, and the caller calls it as a quad, it is downgraded to a triple because of the fall. If the caller calls it as a triple, it gets downgraded to a double, etc.

I don't like that suggestion. Jumps with falls are often downgraded because they were in fact severely underrotated. If we're going to give any points at all for jumps with falls, then falling because the landing foot wasn't even close to rotated backward should be penalized more severely than a fall because the skater did land on the back outside edge with full rotation but couldn't hold the position. It would also mean that falls on single axels would earn no points for the element, only negative points for the fall.

The problem with the current penalties on falls is that it is an extremely small percentage of the total score.

This is true, at the higher levels. The higher the average scores at that level, the lower percentage that the 1.0 fall deduction represents. For lower skill levels, it represents a larger percentage. Looking only at jumps, singles, doubles, and downgraded triples earn less than 1.0 for -3 GOE, so those jumps already end up with negative points considering the fall deduction as well. Where as harder triples and quads still end up with more points than a good double or easy triple, respectively.

So I think it would be disastrous for all but the top senior skaters to raise the fall deduction to more than 1.0 across the board. What might make sense would be to factor the fall deductions according to the type of event (e.g., 3.0 for senior men, 1.0 for novice ladies, 0.5 below novice) although that's still more punitive to the lower-scoring than to the higher-scoring skaters in the same event.

I think the first place to start might be redefining the negative GOEs as a percentage of the base value of the element, and making them more punitive in general especially for the hardest elements. E.g., -1 could subtract 20 or 25% of the base value; -2 could subtract 30, 40, or 50%; -3 could subtract 50, 60, or 70-75%. That's more or less how it works with some of the lower-scoring elements, but quads still get to well over than half their base value even with -3 GOE. So let's reward quads highly if they're successful but much less than currently if they're severely flawed.

So I would like to see bigger penalties for falls, and either increasing penalties for successive falls, or a way to change (or better defined) the PCS system when there are multiple falls in a performance.

Something like that could work. E.g., 1.0 deduction for the first fall, 1.5 or 2.0 for the second fall, 2.0 or 3.0 for the third, etc. Again, these flat fall deductions will hurt the lower-scoring levels and lower-scoring skaters at any given level more severely than the top skaters. Any flat deduction that will take a significant bite out of top senior men's scores will unnecessarily devastate a weaker junior (or lower) who only does double jumps, which is why I would tie any increased penalties more closely to the base values and component scores that that skater earns.

How about a combination of the following:

1) More severe GOE penalties as outlined above so that high-value jumps lose more than half their value with -3 GOE

2) Slightly increased fall deductions for subsequent falls (1.0, 1.5, 2.0, etc.)

3) Explicit instructions to judges to reduce the Performance/Execution score for each fall, even if it's just "at least" 0.25 per fall (this will be factored, double in the freeskate than the SP) and additional guidelines to reflect falls in the other program components where applicable. As is, some judges do and some don't -- if it's in the written rules, we'd see it happen more. But what they're reducing from is still at each judge's discretion.

This would be in addition to the standard fall deduction.

Again, we want the penalty to be higher numerically for skaters who are starting out with PCS in the 80s than for those starting in the 40s. How can rules or guidelines be written to achieve that goal?

Yes, falls should be penalized more harshly. You fall on an element, you should get no credit for that element - it really doesn't have to be more complicated than that.

That's an easy solution for solo jumps.

For jump combinations and sequences, already if you fall on the first jump you've lost the opportunity to do the subsequent one(s). But what if you fall on the last jump? Would it make sense to give base points for the first jump only and then take the -3 GOE off that base? Especially in long programs where there was no requirement to do the second jump anyway.

What about falls on spins or step sequences or other pair and dance elements? Should a fall at the end of a complicated element invalidate the whole element? Right now, a fall at the beginning of a spin does invalidate it, but if the skater has already done enough of the spin to get credit for a legal element, they get whatever points they deserve for what they actually did, minus 3 points of GOE. For step sequences, even a fall at the beginning or middle doesn't preclude them from getting credit for the rest of the sequence completed after they get up. Is there really any reason to change those rules?

[Skittl1321]

Yes, falls should be penalized more harshly. You fall on an element, you should get no credit for that element - it really doesn't have to be more complicated than that.

Wow, we're gonna see a bunch of doubles then- that will make the sport super exciting!

[gkelly]

Thinking through my issues a little more, here's my revised proposal:

1) Increase the -GOE penalty on the highest value jumps as I mentioned, so that -3 takes away at least half the base value

2) Get rid of the flat 1.0 fall deduction. There will still be a mandatory fall deduction taken by the Technical Controller or the Referee, but it will be scaled to the level of competition and to the number of falls. I'm going to propose 3% of the Total Segment Score for a first fall, 5% for a second fall, 7% for a third, etc. Actual percentages subject to negotiation after running simulations with various short and long programs at various skill levels.

3) Write some guidelines to judges about how to penalize falls and other disruptions in each of the program components. These penalties will be discretionary in that not all falls are equally disruptive and not all viewers (on the judging panel or otherwise) will find the same fall equally disruptive. But there will be official support for penalizing these errors in the various components as appropriate in the estimation of each judge.

[Ziggy]

1. Someone scores high with fall on difficult jumps

2. People complain

3. Falls are penalized more

4. Less people try hard jumps

5. Competitions are won by conservative skates

6. People complain

7. Jump values are raised again

8. go back to 1

The circle of life

This. Just because there is a result that you don't like doesn't mean that the system should get changed upside down. Some people seem to be very short-sighted and don't think about the consequences of what they are proposing.

We got the values of quads raised to IMO ridiculous values because everybody was whining about the lack of quads. And thanks to this, we now have Reynolds winning 4CC and placing 5th at Worlds even though his jump technique is absolutely atrocious and he pretty much never rotates any of his most difficult jumps fully and two-foots them quite often as well. So well done, whiners!

I personally think that quality should be emphasised more than difficulty but there's more to that than falls alone. Also, I think that PCS should factor less in the SP scores. At the moment SP scores count a bit too much in the overall result, IMO and SP is meant to be the technical program after all.

I understand why people might want falls to be deducted more but you have to be extremely careful not to overdo it.

My proposal:

Fall on a scoring element : no penalty but no points earned for that element (obviously some fine tuning would be needed for falls on the second jump in a sequence/combination)

Fall on a non-scoring element : -3 (or more) points

Your proposal would result in nobody attempting jumps that aren't super consistent for them. Why risk a fall on a quad and zero points when you can do a triple jump worth several points instead?

It could also result in a skater giving an incredible performance but losing the gold/other medal because of some freak error that shouldn't really matter much overall.

All falls should not be treated in the same away. Sometimes they are messy and disruptive, sometimes skaters bounce back up straight away and continue as if nothing has happened. If somebody falls 3 times but still presents the program brilliantly, why should they get deducted on the PE for example?