Tales of Woe, Volume I, Issue 7

Contents this Issue:	Letters to the Editor

Words of Wisdom: The Changing Face of SPARF, III

Pontification: Steve Finally Learns!

Forum: Choosing Substitutes

Open Floor Policy

Mistake of the Week: Three Mobiles

Random thoughts for new managers

Section A: LETTERS TO THE EDITOR

From: Kevin Pankhurst (NYHV's; E-mail address deleted)

I see that Mel has several great things planned for the future although I feel there are several small additions that could be added to the list he issued. As well as stating in the scouting report where the sub entered the field you could state what time in the match he came on. This could tie in fairly nicely with information on when the players were injured.

What would be nice is if you state in your lineup is a list of positions the 2 subs should fill (in order of priority) if more than 2 players get injured. e.g.

FILL RKM RKR ROV LFP FF RFP LHF CHF RHF C RW LW LHB CHB RHB LBP RBP FB

[Editor's Note: Actually, there is a system for this: the 2F statement after the "end". Write back if you have questions.]

Therefore if say the RKM, the LHF and the C players get injured then the 2 subs will fill the RKM and the LHF positions (as you consider them the most important). If you then find that another player gets injured and his position is more important than one a sub is filling the sub will move to that position. Therefore in our example, say the LFP player is injured, the player at LHF will move to fill the LFP position instead. This would make sure that important positions (like the mobiles) are not left vacant. This could be expanded so that each sub has a list of priority which will correspond to him (therefore a player better in defence will be more likely to fill a defensive position, etc).

Lastly, for those interested, I put foward my suggestion to Mel (that was printed in issue 6). I did not keep a copy of his comment, but as far as I remember (correct me if I'm wrong Mel), he said that he would certainly consider this suggestion. I suspect that many people would consider it a very fair way of calculating random orders, especially if players are having some problems with their mail (like the poor Tasmanian Devils) or if they are planning to go on holiday. Most people have players that they are adding 4 fatigue to each week and just because they are unable to submit an order for his lineup it does not mean that he should lose most of those players. I myself have 22 regular players and 12 players preparing for next season (I am aware that this was featured as the Mistake of the week in issue 3, but it seems to be working so far as I am keeping down my injuries by having low fatigue) and I would be disappointed if some of those 12 players get injured. Naturally, who says that that hazardous substitution method cannot be kept for when players submit incorrect lineups?

Section B: WORDS OF WISDOM

From: labc-3ad@WEB.berkeley.edu (Mel Nicholson)

Before I start with this week's topic, I'd like to respond to last week's "Letters" section:

  1. I'd like to see a new random order generator (NYH)

    So would I. Currently there is the need for some invariant criterion by which to exculde players in choosing which does not involve reference to other players. The way the lineup maker works is to make a list of people who fit certain criterion, then lift one criterion each time the list is exhausted until all positions are filled (or forfiet if the last criterion are lifted)

    The current criteria are:

    		Not playing (unliftable)
    		Not Recovering
    		Fatigue < 50
    		Fatigue < 25 (just added this one)
    		Fatigue < 10 (this one too)
    

    They get lifted from the bottom up. If there are any alterations you'd like to suggest, feel free.

  2. Why not damp the system more? (BBE) Should the best teams be winning by 200+ to 6? (WPD)

    The system is damped somewhat. I should note in its defense that most of the bizzare injury problems are fatigue related, which can be damped by not letting your club get to the deep end of double digits. As for the best teams making ridiculous blowouts, a lot of them have been the recipient of the "thrown game" where teams will hide their stars (to avoid injury) and *LET* themselves be blown.

    [Editor's Righteous Indignation: Yeah, well, I went all-out to blast the Redwoods and scored a season-high TWENTY-SIX. Of course, I injured a mobile in the match, but the blowout exists whether I tank or not. So here's one back atcha, Mel: in the UEFL and EEFL (soccer leagues), there is a penalty for "tanking" a match. What is my incentive to not tank a match in SPARF?]

    In any case, I expect this will be a lot less of an issue next season after the league are tiered because the contestants will be closer matched in skill.

  3. What skills are best for CAUSING injuries? (PAAG)

    This brings me to my real topic: injuries. So let me get rid of this Q & A format to save lines...

Injuries get checked for almost constanly during the game, but in most situations the chance of getting on is so close to zero that it doesn't happen. The times injuries DO happen usually involve a defender doing something in a hurry (like tackling someone) or when one of those long, high passes come down and eight thousand people are sitting there just waiting for it (plus that one guy who shows up at a full tilt run to jump on your fullback shoulder's at the last minute). The exact mechanics of how this gets taken care of are far to messy to go in to, but I've found a pretty good formula for calculating the chance someone will get hurt, based on a test version of the program I ran a long time ago, which instead a rolling for a player, just recorded what the roll was. The net result was that the chance of getting injured was roughly:
	P(injury) = 1 - (K^(1+(fatigue/4+age/2)))

Where k is about 107/108. Note that this doesn't take into account the position your player is at, or who he is up against. This is a general average based on a large number of positions and circumstances. For example, in a game you're being blown out in, your Full Forward will probably not get the ball much, and have a lower chance of getting injured, and the Fullback had be have his insurance paid up. Likewise, the mobiles are almost always more prone to getting hurt. This formula tend to get off a bit when dealing with the age/fatigue levels over about 2/25 (where this slightly overestimates the chance of injury), but for most it works well.

But back to the question, what skills help injuries? I don't know, really. I used to think that a high defense would help, because you could tackle people better, but now I'm not so sure, as if your defense is good enough, the should never be open enough for someone to try to pass the ball in the first place. I suppose injuries are most likely when the player is good enough to get the ball, but not so good as to avoid being tackled.

Having said that, I should note that this is a TEENY TINY little factor compared to the age and fatigue of the player. In fact, I'd guess that if you wanted to try a linear weighting of the importance of age and fatigue against all other factors (including position), you'd find that the latter are less than 3% responsible for an injury. Fatigue is the big factor you can control --- everything else is mostly superstition.

Section C: PONTIFICATION

It's time to return to the original Pontification of the whole season and take a look at what went wrong and what we lucked into.

I will not go into grotesque detail of how I used to manage my team or why some of my strategies were slightly more constructive than using, say, a Ouija board. Here are some things, though, that can provide you with what I consider are general guidelines for future success.

  1. Split your coaches into two basic groups: starters and developers. I have chosen to have five coaches slotted every week to making my "starters" (a term defined below) consistently better without cranking the fatigue too high. The number of players you are willing to field each week should hover in the neighborhood of 21-24 (see previous issues), so five coaches makes for nice small groups (max 6) and can add some good skill (or take away little for those at -2 or -4 Fatigue). This leaves two coaches to develop players for next season.

  2. Use one of the development slots for training future stars. Make one of the two remaining coaches (in my case; your mileage may vary) a tiny slot of two or three players and crank some promising youngsters by +4 Fatigue every week. It should be obvious, though, that regardless of the SIZE of these groups, these are guys you don't want to play and should be cranking at +4 every week since Fatigue gets reset at season's end.

    This is how, for example, I theorize that Allen Wrench of the Redwoods was built by a manager with FAAAAAAAR more insight than I had last season.

  3. Among your starters, develop at least 4 players that can play mobile without too many histrionics. See Section E for details.

  4. Your starters should be put into slots that allow like positional players to be coached together. For example, many people have learned the "de- fensive crank" (Stig of Bodoe is the best I've seen); try to build groups that allow these guys to get cranked together so you don't end up killing your forward's Kick or cranking his Defense unnecessarily.

  5. Never start one of your developmental players unless it is the first few weeks of the season and they are at low Fatigue. WPD has bought no less than 15 players since the league's inception: they are out there if you are willing to play a lesser goober in exchange for leaving your devel- opmental squad pristine and injury-free.

    There are exceptions, of course. When Wimbledon (for example: old-timers may remember CSUA as the Injury King) lost 11 players in one week, well, an ounce of prevention was what was necessary, because there's only a few ounces of cure out there in the way of spare gunk, not a whole pound.

  6. Which brings us to Fatigue in general. I still live and die (and bleed and moan) with the "Bogus 4" rule, which says that a team of reasonably-aged Fatigue 4 players will expect one minor injury per week. I have played slightly over this guideline this season and experienced slightly higher injury rates. (The interested reader is refered to Section B where Mel explains the REAL injury algorithm, but learning this is not Fun)

    This means that keeping your players at fatigue 0-2 is being overly cautious. But far more likely are those teams trotting double-digit Fatiguers out week after week, and IN MY OPINION (this is Pontification, remember), that is an UNREASONABLE RISK.

  7. Scout your opponent as much as possible, including trading scouting reports with fellow divisional managers and subscribing to KWPD. I have honestly (honest!) won two more matches this season than I would have without cable, because I have learned my opponents and (thanks to Mel) finally understand a bit about how to read a scouting report.

  8. Remember the basic guideline for "offensive contribution":
    	X = (2 * (Scr - Opp(Def))) + (Mark - Opp(Mark))
    	(where Opp stands for the skill of the opposing player:
    	 LFP vs. Opp(RBP) for example)
    

    If X is positive, the offensive player gets scoring opportunities (or creates them in the case of midfielders); negative means his contribution is severely limited.

    Among other things, it means it is NOT sufficient to have a Defense of 40 to shut down a star FF. If your defender's Mark is 5, then the FF can still create plenty of problems (which seems rather realistic when you think about it). It also means that Scramble is more important that Mark for YOUR forwards, and it means that your mobiles have to be serious about everything since X is calculated both ways for you and your opponent.

    Now, let's see if I can parley this into second place...

Section D: FORUM

Last issue (you can't remember back that far, can you?), we asked what sort of substitution algorithms people use. Nobody cared, so I get to Pontificate again.

I usually pick two of my extremely-high fatigue older players I care very little about, because they are acceptibly mediocre and face a decreased risk of injury for being subs. HOWEVER...

I have had two mobiles injured in successive weeks. This means that NEXT season I will take my own advice, built a fourth mobile, and use him as my substitute in the week he is resting (presumbly at higher fatigue). I will use him at IC2 and specify him (via the 2F option) to be my mobile sub as well as whatever I think is more important of offense or defense.

So, anyone want to talk about anything else?

Section E: MISTAKE OF THE WEEK

This week's Mistake comes from Yours Truly, who has been bitten in consecutive weeks by the same bloody Mistake.

A new team starts with those five wonderful straight-30 Age 3 players just begging to be your mobiles. So I made three mobiles, one an offensive specialist, and traded one after he got injured. Fine, right?

Extra-super-duper-triple-backflip-with-a-cherry-on-top WRONG. You MUST have more than three exceptional mobile-calibre players on your squad. Yes, you do. Don't argue with me, son. Just listen.

If you have three mobiles that are clearly superior to the rest of your team, what do you do when one gets injured? Call in the Fatigue 52 player you are grooming for next season? Pull in the great Scramble/Kick foward with the Defense of 7? Have you ever tried trading for someone else's mobile? (Editor's Note: Don't waste your time)

You need at least four mobiles. You just do. Put them in a training rotation that allows one to sub in every week and specify him (via the 2F lineup option) as your mobile sub. I completely waxed the Gladiators with Edward Teller at ROV. Was Teller good? On the contrary: he was 10 points worse at everything than the starting ROV for the Gladiators. But when the Glad's ROV got injured, that left Elmo the Raging Queen at AVG and LOUSY to come in, and Teller got a WPD-record 7 goals.

Build another mobile. Do it now.

Section F: RANDOM THOUGHTS FOR NEW MANAGERS

Okay, you've wandered through 2/3 of a season and you finally have the hang of it. You've cut the injuries down, you have won a few (or not), and generally now feel pretty good about either the team you are fielding or your chances of improving next season. What should you be looking out for to prepare for next season?

Well, first of all, you don't get 10 brand new Age 0 players; in fact, we returning managers got significantly less (an unlimited supply of really atrocious players or expensive bad ones). So if you were thinking of cutting some of your more gooberish players, I would suggest you hold onto them for next season. This year's injury risk becomes next year's starting LW.

Second, it is my opinion that any time spent actually CALCULATING how much your players lose due to aging is time spent foolishly. In Other Games (tm), this is crucial in planning. In SPARF, general terms are just fine. Yes, your mobiles will be less effective. Big deal. Like you could do anything about it.

Finally, start figuring out what to do with those players you have been cranking all season. I cranked my losers very evenly last season (largely because the skill might as well have been labelled Harry, Fran, Kukla, and Ollie for all I knew of the game mechanics), and I have changed my mind into creating offensive and defensive specialists.