|
MLB Lead Animator
OVR: 13
Join Date: Feb 2003
Posts: 290
|
Re: AI Follow-up Questions...ie The Boys are back in town....
As promised, we're having a followup Q&A session on AI/gameplay now that the game is shipped. I'll answer 10-20 questions on pitching, batting, gameplay balance, player attributes/data, and certain gameplay additions (pitcher-batter analysis, gamelog, new guess pitch, other pitching/batting changes). I'm not qualified to answer about other areas such as graphics, visual-only items, fielding, running, trades, simulated games, etc... I also can't get into future plans, development techniques, etc..
We'll give a few days to gather questions depending on the volume. Here are a few notes ahead of time:
Multiplatform
*) In terms of pitching, batting, and gameplay, almost all improvements have been made equally to every platform (PS3/PS2/PSP). PSP's smaller resolution does require a few minor tuning changes. PS3 has more animation variety to visualize equivalent results. Graphics on PS3 are also visually superior and can influence the experience of gameplay quality. Players who prioritize gameplay won't notice as much difference as players who prefer graphics.
Classic vs. Meter
*) Classic and Meter pitching produce fairly consistent results when used by a skilled player. They're both retuned every year in detail like most of pitching/batting. Some users will be better at one than the other (question of strategy vs. execution skill). Holding X down longer on classic will trade some accuracy for power (velocity/break/fatigue).
Difficulty:
*) Allstar difficulty (normal) is best for most non-beginners. The game defaults to Veteran (easy) for beginners. Veteran is also good for players who aren't looking for Allstar-level challenge (i.e. losing about half the time). To play HoF (hard) and consistently win even half the time, you have to be better than 90% of players. In other words, HoF is for the best player out of 10, and they'll typically win at least 1/2 of 162 games with an average team. On OS, that might be the best out of five players. If you need handicapped sliders just to play on HOF, you'd play a more realistic game on Allstar. The difference between Allstar and HoF is difficulty, not AI intelligence. The increase in difficulty is split between batting and pitching, so each difference may seem minor even though it's enough to eliminate most users from qualifying. Skill/ability in baseball is revealed over multiple games, and as previously mentioned don't be surprised to win a few games on HoF and lose a few on Veteran. Even AA/AAA teams can beat MLB teams on occasion.
Guess/Noguess Pitch
*) The game is designed to play well whether you never use guess pitch or always use guess pitch. You can also play guess type only or guess zone only. Again, some users will be better at one than the other (emphasis on strategy vs. pure execution skill). Guess pitch tends to help beginners and those who are stronger at strategy than reaction, but it's often balanced for a skilled user. Note that Veteran is an exception since the AI adapts slower to your patterns/strategy than on Allstar/HoF - otherwise real beginners wouldn't stand a chance. So "fake beginners" (experts playing on Veteran) therefore may conclude that guess pitch is easier when that's not necessarily true on Allstar/HoF. The AI on Veteran thinks like an intermediate-beginner. Expert users may find guessing only type or only zone (or not at all) to be more optimal, depending on the count and pitcher's mix of pitches.
*) The modes of guess pitch are: New, Classic 1/4, Classic 1/6, No Feedback, and Off. Within guess pitch you can guess both type on zone, only type or zone, or neither. "New" guess pitch tells you if you guessed location correct, without showing you the location. Classic 1/4 shows you the location. Classic 1/6 is the same but your guesses cover 1/6 of the strike zone. No feedback never tells you if you guessed correct. The bonuses/penalties are carefully tuned differently for every combination of usage and result (yes every combo), and each detail has been retuned from last year (mainly for realism/balance/challenge). Note that using the L stick you can always manually guess pitch (the old fashioned way in real life) but the full explanation is too large for the space I have.
Hitting Tips
*) Be patient at bat, and distinguish between fastballs and offspeeds (timing-wise, and be picky on offspeed before strike 2). Before strike 2, only swing at good pitches. If the current pitch isn't good - even if it's a strike - chances are a future pitch is better. You have to swing a little earlier on inside pitches and later on outside pitches to put them in play - it'll get harder to time quality hits on extremes. Swinging slightly early relative to the pitch can get you a little more power, a little late can give you a better look at the pitch (contact vs. power style).
*) When you've gotten the hang of facing pitches down the middle, learn to use the L stick to cover the plate better. This is more important on higher difficulties and is not necessary for beginners. Even though pitches down the middle are the best to time/cover, the pitcher is trying to pitch around you. If you cover 2/3 of the SZ with the L stick (ignoring 3 other zones), you can improve your pitch recognition (ball/strike location, even guess pitch type) and narrow down your timing window (if looking in/out). This is manual guess pitch, which has been around a hundred years in baseball.
*) You have to stay ahead of the AI so it won't catch on (easier on Vet). Note that if you look for an inside pitch, you've filtered out some offspeeds (which tend to be low/away) and will improve your timing potential (both due to smaller window of coverage and fastball filtering). On the other hand, looking away will tend to give you more time to look at pitches. Looking low will qualify many pitches but have less SLG potential, while looking high gives fewer pitches but with higher SLG potential. If you don't have a plan or have trouble working one side, it may be better to leave the L stick centered. I'll post more details later but these are a few examples.
Demo results
*) I saw countryboy's 100 demo game result and was impressed at the numbers, considering it was HoF (hard). Note that those numbers will significantly differ from MLB averages due to the setup. Beckett and Francis are top pitchers, with Beckett having a fairly specific style. Fenway Park has a fairly unusual park factor. The Rockies and Red Sox are not typical teams offense/defense. Every user will have their own tendencies. HOF is hard, and like all users countryboy won't represent the average user (i.e. the average family has 2.2 children, but no family actually has 2.2 children).
Sliders vs. Skill
*) The contact/power sliders can be a substitute for skills that are normally gained through the learning curve. It could be a valid substitute depending on your priorities and pace of improvement, and I'd give at least 5-10 games to improve first without changes. Stats won't tend to stabalize until 10-20 games. Note the factors described in the "demo results" comment. Sliders can also be a way to neutralize your own tendencies. While that could be what you want, it might not be if your tendencies are a result of specific strategies (pull hitting vs. opposite field hitting) or difficulty level (particularly Veteran or HoF). Certain hitting sliders have natural but unwanted side effects when moved past a few ticks. To maintain realistic play, I'd recommend focusing on the pitching BB/9 and timing window sliders to tune for individual skill (varies from person to person).
Power swing
*) Power swing is ideal for 2-0 and 3-0, where you want to either hit it hard or not put it into play at all. Many batters aren't even allowed to swing on some counts (though less true in MLB than lower leagues), but when they are allowed that's a good mentality ("hit it hard"). You don't want to put the ball weakly into play when you're ahead like that, when you could easily get on base on ball 4 or potentially see 1-3 easy pitches instead. On a power swing you get a moderate power bonus, you are fully committed (cannot check your swing), your timing window is a little harder (slightly more fouls), and it's harder to make adjustments (chasing pitches outside of the strike zone or your coverage). If you don't know what you're doing, you're much better off not power swinging in the long run.
*) Ahead counts (such as 2-0/3-0) are also ideal because a lot of pitches are fastballs deep in the strike zone (predictable timing, location). Note that from 2-0/3-0 you get at least 3 chances from a position of advantage (misses/fouls are low-cost), so you don't want to swing at just any pitch (regardless of if it's a strike). Other clutch situations make sense for advanced users, or if you're dominating the pitcher. Power swinging obviously won't help if you end up chopping the ball or popping it up (or fouling it), and it tends to hurt AVG/Ks/BBs in return for SLG/RBIs in the long run.
Pull/Push Hitting
*) Player strategy has a big influence on results. For example, batting opposite field like Derek Jeter or Ichiro and gaining contact at the cost of power. These players will tend to do that naturally, but you can choose to shift every player in that direction to some degree by choice. By swinging later on pitches, you gain pitch recognition (type and location) at the cost of some power (pushing vs. pulling). Power strategies work the other way and require excellent pitch recognition without hesitation. The most powerful swing requires you to be slightly early (P for perfect in pitcher/batter analysis) - obviously trajectory plays a dominant factor. In the game, contact strategies tend to be optimal for contact hitters, and power strategies tend to be optimal for power hitters. But it's a minor factor that only matters in HoF level play, because there are much more important skills to focus on like taking pitches well.
*) More generally, players who try biased strategies will get biased results. Your hitting results will only be realistic when considering what you've done. Most beginners hit for contact and don't realize it. In real life some normal batters will become contact/push hitters when overmatched by an ace for similar reasons, many times by choice.
Statistical Streaks
*) There are 30 teams in the MLB, but perhaps a million people who will play the game. Every person will get different results and different streaks. Let's take the Red Sox in real life and their first 5 games of 2007 (April 2nd-7th). They hit 1 HR out of 5 games. This isn't enough to draw any broad conclusions about MLB or the Red Sox's power. If I pick another team, there will be another stat I can focus on. If 1 team out of 30 had this kind of streak, imagine the variety in a million people. I've found that it takes 10-20 games before most stats will stabalize, and in multiple runs there will eventually be a set of 10-20 games that strays from the trend. You can see this in a good/bad month for a team. In internal development, I never hear anything from the 90% of people getting expected results, just the 10% who get the unexpected.
Other
*) The game fully supports Dualshock 3.
*) Repeated steals/bunts/etc increase the alertness of defense. Against a high steal threat, you want to focus more on fastballs. Against anticipated steals, pitchouts can be very effective (the value of an out is very high compared to the cost of one ball).
*) Fastballs are fundamental because of their accuracy/reliability. A properly thrown curveball is harder to hit, but a badly thrown curveball can be taken for a ball or slammed hard for a hit (undoing the benefits of several good curveballs). More pitching tips in the future.
*) High velocity pitchers don't have to worry about location as much, and have a shorter learning curve. Therefore beginners will tend to find these types of pitchers "easier" to use, when a skilled player wouldn't necessarily. High control/movement pitchers focus more on strategy (pitch type/location selection) and hitting location. While they have a longer learning curve and can't afford to miss as much, they also have greater situational control. An example of situational control is whether you walk a guy without wanting to vs. walking a guy due to a calculated risk (not intentional but not completely unintentional). At 95-100 mph, location is not as important, although it's certainly not optimal to throw pitches down the middle in the long run especially when batters catch on. At below 88mph, missing your location can be much more costly and you rely more on changing speeds, location, and pitch movement. I.e. confusing the batter rather than blowing him away.
*) The API pitches differently to every batter, with every pitcher, and for every situation and count. It considers factors that casual players wouldn't appreciate, but are necessary for a realistic game. To the most casual user, I don't expect them to see any improvement on the API - but it serves them a good purpose nonetheless. The API has no hidden knowledge of any kind. It relies on information available to the user and applies that to its strategy. There is no inherent bonus for listening to the API, unless your sense of strategy and situation is weaker than the API. It's defaulted off on Allstar/HoF but you can force it to be on. If you really are of Veteran (easy) skill, you'll definitely learn something from the API. The most important detail is the precision of working strike % by count/type/zone/etc.
|
Operation Sports
03-17-2008, 10:24 PM
#1 (permalink)
). It really is a special game, and I can't believe how much I am playing it and enjoying every game. You guys should all be proud, you've churned out one FANTASTIC video game.
Linear Mode
