Table Difficulty levels based upon scoring statistics

[Vincent]

Hi everyone,

I thought it would be quite instructive to try to classify the tables according to their difficulty levels.

As a guy (or girl) named "wednesday" said here http://forum.zenstudios.com/showpost...1&postcount=36 ,

"As far as scores go, everything is relative. If it's easy for you, then it's easy for everyone."

But if we say that, then we can securely affirm that all tables are of the same difficulty, because we compete against the same people. However we know this is not true. When we play a table, we think "oh, with this one, it's hard to prevent the ball from draining", or "on that one, it's so easy to get 1 million". So that's it, we have the 2 criteria that determine the difficulty of a table :

1. Ball longevity (including Extra Balls)
2. Scoring

As the first one is quite difficult to apprehend in a scientific and objective way, I concentrated on the second one : scoring. So I patiently browsed the leaderboards and took notes in order to determine, for each table, the percentage of players who were able to reach 100 000 000.

Why 100M ? Because on many tables it's a reasonable goal to achieve. See this thread http://forum.zenstudios.com/showthread.php?t=3028

I also calculated 300M rates but it only confirmed the 100M statistics, and it of course looses precision. However I will often give it for information.

When a table has a big 300M rate compared to its 100M (keeping the ratio), it can either mean that the table has a lot of "secrets", things you have to figure out before you can score quickly and efficiently, or that it is very popular among good players.

So I classified the table into 5 categories. I didn't include Paranormal, just because it's a new table and statistics don't reflect the true difficulty level yet. In my opinion, it will be classified among the "Very Easy" tables, because the scoring is very cheap and the ball is easy to keep thanks to the nice outlanes, the magna-save, or the easy extra-balls.



I. "Very easy" tables : more than 20 % of players manage to reach 100M ; 6 % reach 300M.

=> Very high scoring.
=> Very high ball longevity.

*Ms. Splosion Man (36 %)

*Sorcerer's Lair (22 %)



II. "Easy" tables : 5 % to 10 % of players manage to reach 100M ; 0.5 % to 1 % reach 300M.

=> High scoring.
=> High ball longevity.

*Mars (10 %)

*Captain America (7 %)

*Spiderman (6 %)

*Fantastic Four (6 %)



III. "Average" tables : about 1 % of players manage to reach 100M ; less than 0.1% reach 300M.

=> Average scoring.
=> Average ball longevity.

*Earth Defense (1.2 %) [300M = 0.05 % ; only 1 billionaire]

*Wolverine (1 %) [300M = 0.12 % ; 35 billionaires]

*Pasha (0.9 %) [300M = 0.05 %]

*Secrets of the Deep (0.7 %) [300M = 0.09 %]

*Biolab (0.7 %) [300M = 0.05 %]



IV. "Hard" tables : less than 0.5 % of players manage to reach 100M ; about 0.05 % reach 300M.


a) Average scoring / Poor ball longevity.

*Rome (0.5 %) [Borderline ; but 300M = 0.03%, and only 3 billionaires ; a lot of people consider it harder than the other 3 "core tables"]

*Excalibur (0.35 %) [300M = 0.03 %]

*Blade (0.28 %) [but 300M = 0.08 %, and 14 billionaires => lots of things to "understand" and then it gets easier]


b) Low scoring / Average ball longevity.

*Nightmare Mansion (0.4 %) [300M = 0.03 %]

*Street Fighter II (0.37 %) [300M = 0.02 % ; no billionaire]

*Rocky & Bullwinkle (0.32 %) [300M = 0.04 % ; this table is an example of excellent scoring scale "re-calibrating", unlike Buccaneer...]



V. "Very Hard" tables : less than 0.2 % of players reach 100M.


=> either very low scoring (classic tables) or low scoring + poor ball longevity (Iron Man)

=> very few people make it to 300M


*Agent (0.13 %) [300M = 7 players ; no billionaire]

*Iron Man (0.08 %) [300M = 2 players ; no billionaire]

*Speed Machine (0.05 %) [300M = 2 players ; no billionaire]

*Buccaneer (0.05 %) [300M = 5 players ; no billionaire]

*Extreme (0.02 %) [300M = 6 players ; 3 billionaires, players of exception]


So as a conclusion, I would say that tables of categories II, III and IV are very well equilibrated. Again it confirms the very high quality of the core pack, whose 4 tables have a balanced difficulty level. On the other hand I believe tables of I and V level difficulties have a little problem. Of course like everyone I love Sorcerer's Lair, but Ms. Splosion Man is really not my cup of tea because I just can't spend so much time on a single table without ever losing the ball, and as for Buccaneer & Co. they have a crappy scoring system and should really be updated... well actually, they were ! and that's the problem : when switching from PFX1 to PFX2 the scores were considerably lowered in order to harmonize the scoring with the new tables, all right Zen but you failed because, now, it's far too low ! Of course that's a little thing and we still love you...


(This post may be updated as future tables are released.)

[lexandro]

Zen already have a major update in the works thats intended to change how the tables are scored for tournaments and as far as I am aware for leaderboards also.

[Vincent]

Zen already have a major update in the works thats intended to change how the tables are scored for tournaments and as far as I am aware for leaderboards also.

I didn't know that. Very good news, thanks !

[Flynn74]

So as a conclusion, I would say that tables of categories II, III and IV are very well equilibrated. Again it confirms the very high quality of the core pack, whose 4 tables have a balanced difficulty level. On the other hand I believe tables of I and V level difficulties have a little problem. Of course like everyone I love Sorcerer's Lair, but Ms. Splosion Man is really not my cup of tea because I just can't spend so much time on a single table without ever losing the ball, and as for Buccaneer & Co. they have a crapy scoring system and should really be updated... well actually, they were ! and that's the problem : when switching from PFX1 to PFX2 the scores were considerably lowered in order to harmonize the scoring with the new tables, all right Zen but you failed because, now, it's far too low ! Of course that's a little thing and we still love you...

Great post, Vincent. Hopefully, they will fix the superscore problem in the 2.5 update next year.

[Vincent]

Great post, Vincent. Hopefully, they will fix the superscore problem in the 2.5 update next year.

Thank you ! Wow that's a long quote lol

[Flynn74]

lol. Yeah, mate, you're right. Fixed ;-)

[Pbdrums001]

Thanks!

Thank you for putting this together. This is very interesting and clears a lot of confusion I've had on some of the tables. This also makes me pretty proud of my 114M score on Excalibur

[Vincent]

lol. Yeah, mate, you're right. Fixed ;-)

Nice

This is very interesting and clears a lot of confusion I've had on some of the tables. This also makes me pretty proud of my 114M score on Excalibur

Thank you, it clarified things for me too : I was quite surprised to see Mars among the easy tables since some targets are so hard to shoot (pyramid), I didn't expect Excalibur to be among the hard tables and I'd never have thought Iron Man was so hard. You can be proud of your Excalibur score. By the way I love that table !

[Hypno74]

Salut Vincent.
Je vais répondre en français, un peu de chauvinisme ça fait pas de mal
Bien joué, c'est un gros travail pour essayer de faire avancer le shmilblik.
Je suis sur PS3 mais nous avons beaucoup de tables communes.

Je suis assez d'accord avec ce système de classement même si j'émettrais quelques réserves.
Les scores ne reflètent pas forcément la dureté des tables même si ils sont élevés (si on omet les failles).
Par exemple, Earth defense, Excalibur ou Mars sont pour moi des tables très difficile en terme de jeu "pur", si on veut faire le tour de toutes les missions.
Blade n'est pas vraiment une table difficile, il est assez aisé de faire plusieurs fois le mode sorcier.
Pour Iron man, imaginons que la notation soit plus élevée : 20 millions par mission réussie et 50 millions pour Ultimo. Les scores seraient plus gigantesques et la table passerait en catégorie Médium.

Il y aurait peut-être d'autres paramètres à mettre dans la balance.
En tout cas, merci pour cette analyse.

[Flynn74]

I appreciate the time and effort taken to put the original post together. However, I think Spider-man should be classed as an easy table. The wizard mode is one of the easiest to get to in the PBFX2 library. As soon as you understand the levelling aspect of the villains, and that it carries forward to future games, it is relatively easy to achieve a very high score. Pretty much agree with everything else, with the possible exception being Ironman.

Zen really need to address this scoring imbalance between the tables in sections I and V. I have less motivation to play the 'very hard tables' because they are so low-scoring. Also, the PBFX1 core tables feel broken to me. I managed to get to Ultimo for the first time today on Iron-man and feel pretty good about my 86m score. If Ironman was slightly higher scoring I would play it a lot more. It's a good, if challenging table and I'm not sure if it deserves to be lumped in with the PBFX1 core pack.

However, the stats are hard to argue against.

[Vincent]

Salut Vincent.
Je vais répondre en français, un peu de chauvinisme ça fait pas de mal

Sympa Par contre si tu me permets je vais continuer en anglais par respect pour les non francophones C'est un peu chiant je sais lol mais bon...

Bien joué, c'est un gros travail pour essayer de faire avancer le shmilblik.

(Well done, that's a big work in order to try to make some progress on the matter.)

Thanks !

Je suis assez d'accord avec ce système de classement même si j'émettrais quelques réserves.

(I rather agree with this ranking system but I would express some restrictions.)

I know what you mean, I pretty much realized that this was in no way a perfect ranking.

Earth defense, Excalibur ou Mars sont pour moi des tables très difficile en terme de jeu "pur", si on veut faire le tour de toutes les missions.

(Earth defense, Excalibur or Mars are in my opinion very difficult tables in terms of pure gameplay, that is to say if you want to complete all the missions.)

That's an excellent point. I completely agree with you, Mars or Earth Defense are very difficult. Also, in Secrets of the Deep, it's impossible to score if you don't know what you're doing. Such parameters should enter into consideration when speaking of the difficulty of tables.

Obviously that was not my point here. Nothing forces a player to make all the missions, the purpose of a pinball game is to score. Discovering is where all the pleasure lies, but yet the final goal is to score !

For example, the quickest way to get extremely high scores in Rome wouldn't be to achieve Wizard Mode, but to keep completing Romulus multiball...

Blade n'est pas vraiment une table difficile, il est assez aisé de faire plusieurs fois le mode sorcier. (Blade isn't really a difficult table, it's quite easy to access the Wizard mode several times.)

It becomes easy once you get to know the table. Acquiring this knowledge is part of the difficulty.

Pour Iron man, imaginons que la notation soit plus élevée : 20 millions par mission réussie et 50 millions pour Ultimo. Les scores seraient plus gigantesques et la table passerait en catégorie Médium.

(On Iron Man, let's imagine if the scoring was higher : 20 millions for each completed mission and 50 millions for Ultimo. The scores would be more huge and the table would be classified as "Medium".)

Of course but you could say that for all the other low-scoring tables ! One of the purposes of my post was to show which tables are not balanced enough ! Like I said, Buccaneer is not supposed to be so tough, it's just the scoring system that is completely messed up !

Il y aurait peut-être d'autres paramètres à mettre dans la balance. (There might be other parameters to take innto consideration in order to make a better judgement.)

It's true that a more complete analysis should take other parameters into consideration. We could count the number of missions or special modes in every table and take their difficulty into consideration, and we could also consider the difficulty of the shots, and the ratio difficulty / reward for multiplayers (in Secrets of the Deep, multiplayers are very hard to get and do not reward much)... People who know the tables by heart will do it if they want. It's a hard work. I know Rome very well but I still don't understand anything about multiplayers...

---------------------------------------------------------------------------------------------------------------------------

I think Spider-man should be classed as an easy table.

Well, it is. You mean very easy ? I think it's very easy yes. But still, why those stats ? Because people don't know how to score very high, and, like I said for Blade, acquiring knowledge of how to score is part of the difficulty, so, in a way, it's really normal that Spiderman is not in the same category as MS. Splosion Man which is not an intelligent table at all and where there is really not much to understand : you don't need to think in order to score, it's a table for children ! Sorcerer's Lair is not like that but you quickly understand how to make hi-scores, and I think Paranormal will fall into that category.

Zen really need to address this scoring imbalance between the tables in sections I and V.

Yes they do !

It's a good, if challenging table and I'm not sure if it deserves to be lumped in with the PBFX1 core pack.

I love Iron Man ! So here again, I understand what you feel. And it shows one thing : that there is a problem with the scoring system : it doesn't match the quality of the table.

However, the stats are hard to argue against.

Well at least they show the real objective difficulty to achieve 100 M. But their problem is that they are mechanical, not intelligent, and therefore they don't make a difference between low-scoring coming from an actual challenging gameplay and low-scoring coming from problematic scoring systems ; in the same way, they don't make justice to the subtility of Secrets of the Deep or to the very challenging aspect of Mars, because they don't make a difference between hi-scoring coming from an easy gameplay and hi-scoring coming from a generous (but not necessarily generous for beginners) scoring system. This is truly not a flawless ranking system.

Maybe I should add stats for a 20M threshold. Problem is that it takes ages to scroll the leaderboards, Zen should do something about it.

[Cloda]

Thanks Vincent for this interesting analysis. There is really so much that can be said but in the end I think I'll try and keep it short and just say that each tables leader board is representative of the skill of the players and their willingness to put time into it - the top 10 scores on each table are usually not in line with the "normal" scores on the table as there is usually a whole lot more effort being put in to get those scores (and exploitation of scoring loopholes ).

So, to get a truer reflection of the difficulty of a table you need to take many other factors into consideration (as some of the guys have said) e.g. scoring rate/hour, how many players have reached the wizard mode (30G achievements) etc. and I don't think that information is available at the moment. It would be great though if Zen brings in some sort of other leaderboard that takes into consideration more aspects of playing a table rather than just ultimate high scores (as have been discussed in this thread) as this will be more representative of the difficulty of a table relative to other tables (if they get the weighting right to compare between tables).

Some of the tables that I will rate difficult for me are (irrespective of the scores that you can get):

- Excalibur (have not been close to Wizard mode)
- Earth Defence (only once got all the medals, but have not defeated the Wizard mode)
- Ironman (just because of Ultimo)
- Rome (have to work hard to reach Wizard mode - only done it a couple of times)
- Blade (have to work hard to reach Wizard mode even though I can reach it quite often)

Here is some tables would have been more difficult in my opinion if Zen just made some adjustments to their scoring:

- Paranormal (only 2 extra balls per ball and no magna save)
- Wolverine (no claw save and only 2 extra balls per ball)
- Sorcerer's Lair (that you hav to complete the missions)

I tend to really like the tables that gives me a solid challenge in a relatively short period of time (this is obviously different for each player depending on their skill level) and these are the tables that I go back to often every couple of months:

- Pasha
- Earth Defence
- Excalibur
- Blade
- Fantastic 4 (once they will finally fix it).
- Rome

Many of the other tables I just avoid because the games are too long to get anywhere. So once I have put in a decent score I will most probably not play the table ever again:

- Paranormal
- Secrets of the Deep
- Wolverine
- Mars
- Ms. Splosion Man

[Womble]

I keep going back to Pasha, even though it's probably the most difficult table for me.

In 80 games of Pasha, my stats:

- zero storybook completions (thus zero wizard modes, no 30G cheevo)
- zero successful magnets games
- ~5% Oasis shot success rate

Earth Defence has some crazy hard shots. Partly because it started life with a different set of physics. But mostly because the designers are sadists. But again, the difficulty (and theme) keeps me coming back.

I find Blade to be a middling game, difficulty wise. Not hard, but I tend to forget how to play the table if I don't play it for more than a week. I am getting old! But when I do play it, I feel like I could go a very long way with it if I had the patience. It helps that you can reattempt the Wizard Mode as often as you like.

I think I've pretty much beaten Wolverine and SotD, scoring wise. Anything more would be a grind.

Excalibur utterly eludes me, difficulty wise. One for a rainy day.

Iron Man just blows. I haven't gotten to Ultimo because it's too tedious to bother. It's not a hard table as such, just an annoying one.

Rome is also tricky, but I think I could get there if I could just develop the patience.

I think Biolab is very hard. I'm missing that part of the brain that other people seem to have. The part that knows what to do on this table. I like the theme and music though.

[Vincent]

I keep going back to Pasha, even though it's probably the most difficult table for me.

In 80 games of Pasha, my stats:

- zero storybook completions (thus zero wizard modes, no 30G cheevo)
- zero successful magnets games
- ~5% Oasis shot success rate

Same stats here .

[Nekro Neko]

I don't understand how people find Pasha harder than Wolverine, I mean, the table is really easy, especially once you get the Magnet timing down. Then again, maybe I say that because I can get to Wizard mode 5 or more times on an average game now.

Wolverine on the other hand I can make it to multiple wizard modes but often don't need to because of Sentinal giving you points like crazy.