Jump to content

Tactical proofs


Recommended Posts

I have posted (here) that I making an experiment with FM2011 game engine.

As you should know, in FM you can find which attributes are important for each role and duties. But importance of attributes are also has sense for different position. For example, AML and STC with similar role and duties has different importance of attributes in reality.

So, which attributes are important for each position regardless to roles and duties?

To define that I made the experiment with team, which I calling "default team". Default team is:

All players has average attributes, generated, and cloned by FMRTE tool. All instructions are similar: running with ball is normal, tackling is normal, passing style is normal, etc

Here is the picture of every player:

CA&PA - 1500

Each attributes (excl Injury Proneness) - 15

Injury Proneness: 1

Years old - 25

Team status: First team

Left foot(Right foot for DL and AML) - 10

No preferred moves

Reputation: 1500

Goalkeeper rating - 1

Height and weight - 180 and 80

Happiness: Happy to stay at the club (75)

Condition: 79%

Morale: Good (15)

Fitness: Needs match practice...(6300)

I have played with this team many seasons and gathered a lot of statistics (part of statistics you can find here).

As the result of experiments I represents here the tactical proofs. This is a list of proved advices that based on importance of attributes for each position (not roles and duties). Here the tactical proofs:

EFFICIENT FINISHING

1.Build you finishing on ability of low through ball to the box.

In terms of type of finishing game engine promote goals from pass to the box (66%) and to one-on-one (23%);long shots scores near 11% of total goals.

In more detailed factor analysis view: 40% from pass to the box without impact other major factors, 18% from crosses, 12% from mistakes of opposite side, 10% from long shots, 10% from FKs, corners and penalties, 7% from tackles on opposite side, 3% from dribbling. So, buy players that able to play passing well, instruct them to create opportunities from passing, and you'll have better success chases than if you'll play attacks from crosses or on long shots.

2. Try to tackle and pressure more on the opposite side if you attacking players are not too bad on tackling, and not tired (has good stamina).

Value pressure more then crosses as it not only create 20% of the goals (by tackling and mistakes), it also make harder your opposite side to control the ball.

3. STC should be able to shoot well from both foots as well as play in the air. Other attackers can play well with single strong foot.

Default team scores most from strong foot (near 58%) and near 31% from weak foot,near 11% from head. But consider that your STC is different, he score 50% from left foot and 50% from right. He should be good in the air as he also makes most of header in default team, but to flick on near player, not to score goals. Do not use AMC,AML and AMR with both strong foots as they shooting primary from single strong foot, and ability to score from both decrease all other attributes very much.

EFFICIENT BALL CONTROL

1.Prevent ball loses by TOTAL accuracy in ball control on EVERY positions.

Game engine promotes most of ball losing from pass intercepting/missing (near 60% of all ball losing for "default team"). Every position is important in passing game. DL,DR, MCL and MCR are important passers as they make more passes then others, AMC, AML, AMR, SC are also important passers as they are major assistants. GK, DCL, DCR makes little less passes, if you have no good passers on this positions, instruct them to pass very short and safe.

2.Be selective in instructing on tackling and heading.

The rest of ball lousing factors for "default team"is: missed shoots, loosed tackles and loosed headers, has near 13% of ball losing for every listed factor. It is not many in general, but.... your STC, DCR and DCL are making near 5 header per game, which is much higher then rest of the team (near 2). Be sure, that they are good on it. Your STC may not score a lot from header but he defiantly able to won many important air balls on the front.

3.Use DMC instead of MC

FM engine makes DMC more specialized on possession control then MC. DMC and MC play similarly in passing and tackling, but DMC make long shots 3 time less often than MC (long shots are inefficient in FM compare to real football) and practically dose not dribble (any way dribbling works better from flanks). So, it is more efficient to have DMC with 20 passing and tackling than MC with 15 passing, tackling, dribbling and long shots.

EFFICIENT FLANK PLAY

1. Use AML and AMR with opposite foot.

If you plaiyng with opposite foot winger like Messi, it can decrease number of dribbling per match (from 3 to 2,5), and increase shooting attempts for wingers (from 1,7 to 2) but shooting attempts form opposite foot are more dangerous and create more goals. Your wingers will score the same amount of goals as STC if they are equal. The rest actions, like passing and assisting, tackling and heading will works the same.

2. Use flanks often, play wide formation in attack AMR+AMC+AML, not AMCR+AMC+AMCL.

Narrow attacking formation(AMCR+AMC+AMCL) not promoting neither passing, no tackling and heading for AMCR and AMCL, but it deffenetly decrise number of dribbling per game for them (from 3 to 1,7) that is why they will create less shooting attempts and goals.

3.Use DL,DR ether as totally supporting ball passers (not crossers). Or use WB instead.

Engine is not promoting DL and DR to make many crosses, dribbling and tackling, so do not instruct them to do that. Instead use their strong side - huge amount of passes per game. Instruct them play like central defenders, without running forwards and crosses but supporting midfield by many passes. But if you want to instruct them to cross, dribble and tackle more, than using WB position is suitable, as this position is promote mentioned attributes more then for DL and DC

SORRY FOR MAY BAD ENGLISH. PLEASE ASK ME IF SOMETHING IS NOT UNDERSTANDABLE.

Link to post
Share on other sites

@furiousuk,

I playing FM longer time than speaking English.))) I realize that my data and tactical proofs can be criticized easily, but they are not controversial, I have proved every my advice by repeating experiment several times. My conclusion based on big amount of data, In my "saved games" I have much more data than I represent here.

Maybe it not covered all aspects of the tactics. And same of advices may looks too obvious, but I put there only what I am 100% confident. That is why it is not Tactical Theorems it is Tactical Proofs.

Link to post
Share on other sites

And same of advices may looks too obvious, but I put there only what I am 100% confident. That is why it is not Tactical Theorems it is Tactical Proofs.

Even if you're 100% confident of something doesn't mean it's true. In fact I myself am 100% confident that everything that you claim isn't true.

Also I would point out that clearly you don't have any idea what the terms theorem and proof actually mean. (Wikipedia will be sufficient in this case.)

You can prove these things only if you have access to the code itself (which you don't). Even with the code I claim that proving anything but the simplest things (which would be almost trivial in any case) will be virtually impossible.

If you wish to find some statistical "truth" then you need to play with every combination of players and every roles and against every combination of players and roles in every formation possible. And repeat the test thousands of times. Still, you couldn't offer those results as Proof, but merely as speculation.

Still, your speculation has some merit - even if it seems mostly trivial (or general at best) to me.

Link to post
Share on other sites

@deserter,

Even if you're 100% confident of something doesn't mean it's true.

of course. This is a forum, everyone able to reply a feedback.

Also I would point out that clearly you don't have any idea what the terms theorem and proof actually mean. (Wikipedia will be sufficient in this case.). You can prove these things only if you have access to the code itself (which you don't).

Not only. I can operate with statistical data, also. Please, see http://en.wikipedia.org/wiki/Statistical_proof

If you wish to find some statistical "truth" then you need to play with every combination of players and every roles and against every combination of players and roles in every formation possible. And repeat the test thousands of times. Still, you couldn't offer those results as Proof, but merely as speculation.

I must agree. My Tactical proofs is valuable only for 4-2-3-1 formation (and mentioned modifications with DMs, WBs and AML/R). I forgot to mention this fact in current post, sorry about that. But I have mentioned this in my previous post.

Link to post
Share on other sites

Well, I assumed you were referring to mathematics because you were comparing theorem against proof. I'm, again, 100% confident that a theorem is something that has proof - that is proved by deducing from axioms and other theorems. That proving by deduction part is called proof and the last part of the proof is called a theorem. As such I don't see much point sniffing at someone using a term theorem and using a term proof over that instead. They are much the same. The thing is that you haven't shown the actual proof. Quite the opposite.

Not that I care too much about the proof anyway when you only have empirical data as a starting point. Not when there's only a thin slice of the whole empirical cake that you have.

Link to post
Share on other sites

I'm, again, 100% confident that a theorem is something that has proof - that is proved by deducing from axioms and other theorems.

I am reading English wikipedia now and understand your point. It seams we have different meaning of theorem in English and in my native language Ukrainian. In Ukrainian we have two types of theorems: 1) so celled "proved theorems"(proofs or theorems in English), and 2) hypothetical theorems (Hypothesis in English. It is theorem that need to be proved).

I trying to say that my advices may be not cover whole aspects of tactics but it more proved unlike other hypothetical unofficial guides.

Not that I care too much about the proof anyway when you only have empirical data as a starting point. Not when there's only a thin slice of the whole empirical cake that you have.

I already understood this point of view, but can't understand why you believe so

Link to post
Share on other sites

great effort. but your tests are done on what type of clubs? Did you have control subject for this experiment? If yes, what club and type of players are they?

Sorry, I speak English not perfect, what did you mean " Did you have control subject for this experiment?"

I played for Blackpool from English PL cause many strong team in that league. Every time I finished right in the middle of the league table. As for type of players - all players in my team are similar(you can see details at main/top post of this conversation), players and teams which played against me was different each game but the same for every single experiment, they mostly represent English PM.

Link to post
Share on other sites

Sorry, I speak English not perfect, what did you mean " Did you have control subject for this experiment?"

A control subject, would be if you the same test with a "normal" team using the same formation. Meaning that, maybe if ran a comparison test say, using a club like Ajax.Or even using one strong team, like Barcelona and one weak team, like Barnet for example. Then compared the results to your initial test. That would be your control subject.

Link to post
Share on other sites

A control subject, would be if you the same test with a "normal" team using the same formation. Meaning that, maybe if ran a comparison test say, using a club like Ajax.Or even using one strong team, like Barcelona and one weak team, like Barnet for example. Then compared the results to your initial test. That would be your control subject.

In the experiment I do used team like Ajax. The team had average players: CA&PA - 1500, Each attributes (excl Injury Proneness) - 15 . As for week team and strong team comparing, I think it will not show nothing new, it can be used only for conformation of the results that I have. It takes a lot of time to do that and results form "average team" are pretty accurate. I am playing now for Dynamo Kyiv with mentioned formation and testing my conclusion. For now I see that results of the experiment and real game with Dynamo Kyiv is very close.

Link to post
Share on other sites

is this test on the 11.3 patch? i have some success playing with wingers in a 442 formation(traditional wingers that don't use opposite foot). i wonder how to make the best out of their abilities.

No, on 11.2.XX. Is ver 11.3 has changings in Match Engine?

As for 442, I am not sure that all my conclusions will work for other tactics than I used in experiment. But I assuming, that with 442 wingers will play more near the brow (wide line) of the field and less in the center. That mean that they will be able to shoot and score less but will cross more, and thats why in your case opposite foot will not work as good as in my, probably. But it is just an assumption.

Link to post
Share on other sites

Switching form AML-AMC-AMR to AMCL-AMC-AMCR promote more crosses and dribbles for DL and DR but decrease it for AMCL and AMCR

So if the players move into the centre they will cross the ball less... At the same time, the players that will occupy the empty space down the flanks will cross more. Great find.

Link to post
Share on other sites

So if the players move into the centre they will cross the ball less... At the same time, the players that will occupy the empty space down the flanks will cross more. Great find.

Let me notice that you told previously following: "I'm not sure that this test that you have run proves anything related to the match engine"

So, now you agree that it is related but dislike that some findings are too obvious?

Yes, they are, but at list it confirmed by my experiment now. Before that you was not sure on 100% that DL and DR will make more crosses if they occupy the empty space down the flanks. :p

Link to post
Share on other sites

Most of the 'proofs' are pretty much common sense to anyone with a vague familiarity with FM. However, as a user of 4-2-3-1 myself, I've tended to use inside forwards, whereas you're suggesting wingers would be more effective. That's an observation worth bearing in mind, although my decision tends to be based on whether the crossing or passing attribute is the stronger.

Link to post
Share on other sites

Most of the 'proofs' are pretty much common sense to anyone with a vague familiarity with FM. However, as a user of 4-2-3-1 myself, I've tended to use inside forwards, whereas you're suggesting wingers would be more effective. That's an observation worth bearing in mind, although my decision tends to be based on whether the crossing or passing attribute is the stronger.

My proofs is based on position not on duties (whether it wingers, or inside forwards or other). The goal is to understand what attributes is important for each position. As for important attributes for duties, it is no needs to make test as you can see important attributes for each duties directly in the game. I tend to use inside forwards too.

My experiment shows that the best way to make AML and AMR score many goals is to have quick AML and AMR inside forwards with opposite preferred food, who "likes to try to beat offside trap" and have tactical skills to do that, and, ofcourse, has good finishing. Kay thing is "likes to try to beat offside trap". If they have it, they will have many one-on-one opportunities, more then STC.

So, I use inside forwards duties too, not wingers. I do recommend to not use wingers-crossers(if you want to cross, use WBL and WBR), instead of wingers use forwards that able to beat offside trap, run away, and score with opposite foot. Inside forwards should do the type of job that STC can't do, who should play well in the air, flick on to your inside forwards and be able to shoot and finish well from both foots.

Link to post
Share on other sites

Interesting. I manage in the lower leagues; I can't buy perfect players and no one has PPMs. What I can try is switching wingers/inside forwards to the opposite side. Actually, I'll give them 'swap places' instructions to observe the effects.

Link to post
Share on other sites

It is important to define do you picking tactic for players, or players for tactic. If you decide to play specific tactic (4-2-3-1) than you have to find players that fit it well (or retrain existing players to play new position that fit them best in your tactics).

Specifically to your situation: it is good idea to switch wingers/inside forward to opposite side only if their attributes and PPMs are not sharpened on crossing and if they very fast and able to beat offside trap. If they not, consider retraining different players to play on this position or buying new. Also, ask them to learn offside trap PPM(I don't know if it is possible, but it is very important, as I sad before). Another one important attributes is strength of opposite foot, it SHOULD BE LOW, winger should be or left footer or right footer, not both. PPM of not using weak foot is also useful here. Remember, that wingers scores primary from strong foot unlike to STC. This is highly important point as if player able to play both foots well than it will decrease many other attributes.

Link to post
Share on other sites

You're clearly not of the lower league world. we have very limited options.

I began by assessing my squad and figured 4-2-3-1 best suited. Then I used match prep to train them up in that formation. I've brought in 3 freebies, one being a versatile left/right fullback and another a left/right winger/inside forward - good for the bench. That's my wage limit gone now. None can cross, none have the mental atts to beat the offside trap, and actually in the few games i experimented with switching wings it was ineffective, so I'm back to keeping it simple.

Basically, players make tactics work - or not.

Link to post
Share on other sites

In the Lower League, I found that the best way to approach it is get the best few players possible, and build a tactic around them. I usually have 3 star players, and I build my squad around those guys. In the Lowe League, it's all about making your stars carry you to glory IMHO. At the top level, it's more about the squad and the team.

Link to post
Share on other sites

@Kawee, I tend to agree with @phnompenhandy,

AI unlike to Human makes opposite instructions only for your key players but not makes for rest of the team. So, your key players may be leveled in many games.

To me it is better to have average team, instead of few stars, but they should has several bright complementary attributes and PPM(s) that can be leveraged by tactical instructions. It is impossible to play well if your players average in everything and has no strong sides at all, whatever tactics you use.

Link to post
Share on other sites

I'm sure there are many ways to succeed. The "stars" way just work for me.

For example, my Basingstoke Town was built around 2 strikers and a central midfielder. They put together 70 goals between the three of them, and powered me to promotion. The other teammates were just average Blue Square players.

I only really build a squad in the League 1.

Link to post
Share on other sites

@Kawee, It may works, but you should know that you can't leverage strength of your stars in many games (and hide weaknesses) due to opposite instructions applied to them. Often, you just exploit their general ability (CA).

In other side, theoretically, as you playing in lower league, your opposites may choosing wrong instructions to your stars as their staff is not skillful. But it is theoretically.

Link to post
Share on other sites

The whole point of having star system is to have unstoppable players. I was able to get my hands on an average League 1/top League 2 players while I was in the Blue Square division, and they were unstoppable. That's the whole point of employing the star system in LLM. Your star players have to be so far above the league average that on average, they will be unstoppable.

Link to post
Share on other sites

Interesting, but it needed to test it to say is it works or not.

Maybe your stars so much more skillful than opposite that opposite can't do anything even with opposite instructions applied to your stars. Interesting how Much Engine really works on this point.

Link to post
Share on other sites

Hello GoodControl. It is some interesting statistics.

1) Can I ask how many goals is the percents in EFFICIENT FINISHING based on?

2) Did you play the matches or watch highlights afterwards?

3) And finally, have you made more detailed stats from corners?

Cheers - David

Link to post
Share on other sites

Default team scores most from strong foot (near 58%) and near 31% from weak foot,near 11% from head. But consider that your STC is different, he score 50% from left foot and 50% from right

4) I know you run 4231, but, by chance you made stastistics on SCR (/ SCL) goals percent right/left/head too?

Link to post
Share on other sites

4) I know you run 4231, but, by chance you made stastistics on SCR (/ SCL) goals percent right/left/head too?

No, I have made it only for AML, AMR, AMC and STC. But notice, that percent of right/left/head goals calculated only for AMR with right foot as a main, and AML with left foot as a main. If it is needed to calculate this for AMR and AMR with opposite foot, I can do it per your request.

As for SCR and SCL: it required to change the formation and many additional tests. So I was not able do make that, it take much time, sorry.

1) Can I ask how many goals is the percents in EFFICIENT FINISHING based on?

I am not able to understand you, as my English is not native language, can you re-formulate the question, please. Percents of what you mean?

2) Did you play the matches or watch highlights afterwards?

Highlights only. I played almost 10 seasons in EPL, so it could take much more time to watch all matches. I know that match engine works differently in match view vs just "holiday option" but it works different in assessing details, as for major factors - it assessing the same way. Hope most of things I tested deals with major engine factors.

3) And finally, have you made more detailed stats from corners?

No. I has only percent of goals from any goal factor including corners. I have studied in more details only goals scored from through ball to the box, as it is major goal factor. What can I say about corners, is that it was instructed as "mixed" in my "default team", and almost all gaols that was scored by DCL and DCR are comes from corners.

My general advice is to make team to play passing and through ball the better than you could, and try to reduce as much of other ways of playing the game as you could (like crossing and long shoots). This conclusion based on fact that most of goals scored form through ball to the box, but also, that most of ball losing happening due to pass mistakes or interceptions.

Link to post
Share on other sites

Thank you for all your considerations. :-)

(@ Question 1, i meant how many goals in total you calculated. but from 10 seasons is answer enough.)

If it is needed to calculate this for AMR and AMR with opposite foot, I can do it per your request.

yes, that would be interesting to me.

Link to post
Share on other sites

I find it quite interesting that the "proofs" presented here calls for playing players on their favourite foot. In the TC, the inside forwards are suggested to play on the opposite foot.

kawee - I'll bet your star strikers had blistering pace. LL defenders are mostly slow and lumbering and can be easily done by nippy forwards.

Relatively yes. Both of them.

Fact of the matter is, if you can get your hands on league 2 players when your team is in a Blue Square division, they will tear up the league.

Link to post
Share on other sites

I find it quite interesting that the "proofs" presented here calls for playing players on their favourite foot. In the TC, the inside forwards are suggested to play on the opposite foot.

Playing people with the opposite foot creates better scoring chances and allows player's to open up their body more when shooting, just like IRL it creates better scoring angles.

Link to post
Share on other sites

@Swahn,

Ok, I'll calculate it for 3 seasons ASAP

@Kawee and @Cleon,

I find it quite interesting that the "proofs" presented here calls for playing players on their favourite foot. In the TC, the inside forwards are suggested to play on the opposite foot.

Playing people with the opposite foot creates better scoring chances and allows player's to open up their body more when shooting, just like IRL it creates better scoring angles.

I recommend not only to play flank attackers with opposite foot but also to has players on this positions that plays only with single strong foot (not both as for STC). But currently it was tested only for non-opposite footers AML and AMR, and I'll calculate it for opposite footers AML and AMR too soon.

Link to post
Share on other sites

Wait, so your findings suggest that we should a left-foot-only player as right-winger, and vice-versa? Wouldn't that make it a bit too easy to counter? Force them out wide, where they have to shoot/cross with their ridiculously weaker foot?

Link to post
Share on other sites

Please, can you reformulate last two questions to simple English? Just unable to understand due to my English.

What you mean "too easy to counter" and "Force them out wide"?

I have found following advantage of opposite foot AML and AMR: it increase shooting attempts from 1,7 to 2 and shooting attempts form opposite foot are more dangerous and create more goals. Your AML and AMR will score the same amount of goals as STC if they are equal.

And following disadvantage: it can decrease number of dribbling per match from 3 to 2,5 (Unfortunately it is so. I can't understand why it happening in FM engine)

As for crossing - my statistics shows that if your AML and AMR will play more through ball to the box than crossing they will create more chances for partners (and if you limit them to cross less, they WILL make more through ball to the box). So I recommend to limit crossing for this positions and use them as attacking or supporting forwards. IMHO, the best position for crossing is WBL and WBR (if you need crossing at all). They makes much more crossing than DL and DR and they do not obligate to make many other creative actions as AML and AMR do, so you can specialize them on crossing (and tackling, unlike to DR and DL).

The rest actions, like passing and assisting, tackling and heading will works the same for any type of AML, AMR.

Link to post
Share on other sites

If left-only footed wingers are better on the right wing, what about full-backs? Have you tried a left full-back on the right wing?

Can't see how that would work unless said player was two footed. in real life I was a natural left footed winger but very one footed. Playing on the right would have meant me coming inside for me to get the ball on my stronger foot in order to create a scoring or crossing opportunity. The advantage of my being on the proper side were that as most people are right footed a right back was on his wrong foot to block me which made it easier for me to dribble past him. edit: one advantage I did have on the right was at corners where it was much easier for me to hit an inswinging ball from the right.

Link to post
Share on other sites

@Swahn and everybody,

I have calculated percent of goals scored from strong foot for opposite foot AML, AMR in addition to classic foot AML and AMR.

So opposite foot winger/inside forward scores 62% of goals from strong foot while classic winger/inside forward 77%.

So, you should consider that opposite foot wingers should not be just Right/Left Foot Only but neither both foot. Maybe no more then 10 points for weak foot I think. All goals from weak foot was simple goals while goals from strong foot are cuts inside goals, one-on-ones and shoots from distance.

The base for calculation is 94 goals from 3 seasons in APL.

Swahn, you has open interesting topic, I'll definitely will pay attention.

@phnompenhandy

I am glade that you noticed the benefits from that. Opposite foot AML and AMR scores much more goals than classic wingers. In addition to simple scoring opportunities they start to score variety of other type of goals, like goals from cut inside and many one-on-ones and shoots from distance. Classic wingers rarely do all that actions.

But classic winger also his advantage if hes opposite foot is very weak. He do not need to use it often but if he will have weak foot it increase many other attributes VERY MUCH. So, it is also an option depending of type of players you has. But ideally I like to use opposite foot inside forwards too. Thats tactics more interesting and not less efficient, IMHO.

Link to post
Share on other sites

But classic winger also his advantage if hes opposite foot is very weak. He do not need to use it often but if he will have weak foot it increase many other attributes VERY MUCH. So, it is also an option depending of type of players you has. But ideally I like to use opposite foot inside forwards too. Thats tactics more interesting and not less efficient, IMHO.

Now this raises a whole new issue. My assman has advised me a few times to train my wingers in the PPM 'avoid using weaker foot' . i didn't really see the point, but maybe if they pick up this PPM it will boost certain attribute increases. Any thoughts?

Link to post
Share on other sites

Now this raises a whole new issue. My assman has advised me a few times to train my wingers in the PPM 'avoid using weaker foot' . i didn't really see the point, but maybe if they pick up this PPM it will boost certain attribute increases. Any thoughts?

Attributes of you winger will be boosted not due too PPM but if his secondary foot is really weak("Right/Left foot only"). As for PPM 'avoid using weaker foot' - it gives advantage to players with one foot only, as he start to use his weak foot less.

IMHO, if you use your AML AMR as a classic winger and he is with one strong foot only, than your assman it totally right.

Link to post
Share on other sites

  • 1 month later...

New tests coming.

During my previous tests with "default team" were all players had all attributes on average level, the team finished in the middle of the league table each time.

Now I going to change attributes and tactical instructions the way that match "tactical proofs'" recommendations listed in this topic. I will create "ideal team" (but CA and PA for all players still will be average) and will play several seasons with it with the goal to see how the recommendations affect league results.

I'll testing usual 4-2-3-1, also two modifications (1)with one DM instead of MC and (2)with WDL/WDRs instead of DL/DR, and two variants of AML and AMR (1)with opposite foot inside forwards and (2)with usual strong foot and weak second food wingers.

Please, tell me if you interesting in this tests and what additional info you would like to take from it.

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...