95% crash rule

[JanK]

Dear Jussi,

During yesterday's TDD I saw that few players (I won't name them) were constantly exploiting the 95% crash rule during qualification jumps. It works fine when you jump very far over HS and fall when trying to do telemark. But it shouldn't be used when you know that your jump won't be very long and you try to qualify to the first round by doing front flip when landing. I don't know opinion of other players but this should be fixed somehow.

Thank you.

[Jussi Koskela]

The max number of crash rule qualifiers can be adjusted in the server settings. Currently it's set to 5.
Changing the limit to 1 or 0 could do the trick.

[JanK]

I would reduce it definitely. But as I said, that's just my opinion.

Thank you!

[Tom2]

The max number of crash rule qualifiers can be adjusted in the server settings. Currently it's set to 5.
Changing the limit to 1 or 0 could do the trick.

I would suggest different approach. Set the minimum number of points for the style marks to be in these 5 people. If someone's abusing the 95% crash rule just by extending distance without landing, it usually means he gets worse style marks. We could set a minimum style points (with max possible 60, minimum could be set around 28-30 or even a bit less) to be even included.* I feel it would be less harsh for such big tournaments.

Reducing the limit to 1 or 0 people would require huge changes in some of current hills (plus it's virtually nothing else than KO qualification system on every hill). Currently, landing on such hills as Lahti HS130, Falun HS134, Courchevel HS135, Trondheim HS138 or Lillehammer HS140, Planica HS240 and some more can turn into an utter nightmare. Even top players (including Tour de DSJ winners) struggle big time with some of these hills, humorous examples of that can be found on my channel:

https://youtu.be/csKZNSDs4Mw (Planica HS240)
https://youtu.be/oTxMbE8wA3w (Lillehammer HS138)



*I obviously could do some experiments on checking that minimum, but maybe Jussi knows best what his algorithm will do when you don't click landing and try to extend the distance, it's a very common technique used in 99.9% cases that Jan mentions.

[Tom2]

In addition, let's give an example even from the latest TDD - Lahti HS130. Link to qualifications available here: https://dsjtournaments.com/tournaments/ ... 54023/qual

TL:DR; This one competition was enough to make these top players fall at least once: Bartek Winczaszek, Paweł Wąchała, Altair Ptakowski and Antek Ludwik.

There were 39 falls out of 144 competitors in these qualifications. I chose 30 of them, ending at the "longest" jump of that round - 134.13m from Sebastian Zubrzak and checked the style marks on them.

Link 1: https://cdn.discordapp.com/attachments/ ... _falls.png
Link 2 (sorted by the highmost style marks - 38 points!): https://cdn.discordapp.com/attachments/ ... _falls.png

From link 1, only first 5 places qualified for the competition. From link 2, you can see there were some even higher rated jumps that definitely would've been enough to be qualified had they been landed. I'm 100% sure all the jumps from Jędrzej Książek (38pts) down to Marzena Siwczyńska (32.5pts) were normal falls with telemark attempts. I also believe that jumps between MPH (30.5pts) and mateusz krol (27pts) were landed fair and square. More doubts on Mateusz Pietrzak 3 (26pts) down to 25.5, and the last three jumps (22.5-20.5) were 100% unfairly extended. There's even the biggest gap (3 points) between 25.5 and 22.5 that might suggest there's some correlation between landing and not landing the jump.

So, what's the conclusion? As you can see, it's not always the case that everyone gets a bad jump and counts on fall to qualify. It really depends on the hill, and obviously the players.

Moreover, if we set the minimum style points border wisely, not only everyone who extends their jump will get eliminated, but there also might be a case when the current limit of 5 people could be easily increased, thus making the competitions fair and square.

[Jussi Koskela]

Ok, I won't change the number of crash rule qualifiers. Taking style points into account or analyzing if the crash was deliberate requires more consideration and changes to both server and client.

Thinking out of the box: very bad (deliberate) crash could also result injury meaning that the player won't be able to compete on the next round.

[Xinitiao]

Thinking out of the box: very bad (deliberate) crash could also result injury meaning that the player won't be able to compete on the next round.

Ooooh, that's a very interesting idea, I'd love to see an option for that in offline It could even i.e. measure forces inflicted on key parts of jumper's body and determine the length of the break based on that.
And it could make the start and end of calendar feature more useful and immersive - if the length of the break would be determined in days.

[Maciejo-96]

Reducing the limit to 1 or 0 people would require huge changes in some of current hills (plus it's virtually nothing else than KO qualification system on every hill). Currently, landing on such hills as Lahti HS130, Falun HS134, Courchevel HS135, Trondheim HS138 or Lillehammer HS140, Planica HS240 and some more can turn into an utter nightmare. Even top players (including Tour de DSJ winners) struggle big time with some of these hills

In that case I think it won't happen without jumping physics and landing physics modifications. Seems more difficult than some players might think.

[Xinitiao]

Reducing the limit to 1 or 0 people would require huge changes in some of current hills (plus it's virtually nothing else than KO qualification system on every hill). Currently, landing on such hills as Lahti HS130, Falun HS134, Courchevel HS135, Trondheim HS138 or Lillehammer HS140, Planica HS240 and some more can turn into an utter nightmare. Even top players (including Tour de DSJ winners) struggle big time with some of these hills

Well, imo that happens mainly because the difference in difficulty of landing telemark vs crouched is way too small so players instead of having a strategic choice if to shorten the jump and land telemark or fly as far as you can and crouch, in DSJ4 you basically always try to land telemark and either land or have a big chance to qualify with 95% rule - landing crouched basically means you don't advance to the next round for sure, while landing telemark is a smaller or bigger gamble.

[Jussi Koskela]

In that case I think it won't happen without jumping physics and landing physics modifications. Seems more difficult than some players might think.

Resolving injuries is doable (in a similar way that I calculate whether the skis will detach), but yes, it requires significant amount of work.

[JanK]

Yes, as Tom2 written...I don't want to affect top players who fall after very long jump with telemark landing. My initial though was just to prevent deliberate crashes. This would be very good change for tournaments.

[Siatko]

Personally (as you can see from Tom's material), I think even this top five 95% system can hurt. It's worth looking at the fact that such falls in strong winds at Planica or Lillehammer are sometimes after 10. So in my opinion, it can hurt such players a lot. Additionally, it is known that the rating is not the most important, but for example the athletes, who have a peak rating, simply won't be able to get into the qualification due to the introduction of the system of promotion of only one person, even though they gave maybe even the third furthest distance in the qualification, and thus they will lose even 70 or 60 points. Therefore, if I had to choose, I would leave the current system and not change it so much.

[Xinitiao]

the athletes, who have a peak rating