MapHaven : Playtesters

The playtesting of a map is exceedingly important. MapHaven requires that all maps available on this site be playtested, simply because it makes for better maps.

The mapmaker must understand that they are too close to their map to judge it entirely properly. Granted, a mapmaker on their own can make a very good map - if they put enough thought into it - but any mapmaker, whether expert or novice, thoughtful or mouse-happy, can make a better map if they get playtester input.

In particular, a mapmaker always has a hard time judging a map's difficulty, and the amount of information they need to give the player so that the player enjoys completing the map. The mapmaker can't help but know where everything is - what the answers are, how the map is *supposed* to be played. They can guess ... and then the playtester will tell them how close they are and which direction they need to change things in.

MapHaven has two playtesting systems aimed at linking mapmakers to playtesters. One is "spur-of-the-moment" playtesting, where the mapmaker puts the map on MapHaven labelled as "Awaiting Playtesting" and hopes that someone comes along and playtests it. The other is "volunteer" playtesting, where the playtester puts their name on MapHaven and mapmakers email them and ask them if they will playtest a particular map.

"Spur-of-the-moment" playtesting works roughly like this:

The mapmaker logs into MapHaven. (They should record their username and password in a safe place; they'll be needed again.) They create a map on MapHaven, by filling out a form with information about the map and then uploading the map, zipped with a readme file. However, when filling out this form, the mapmaker does not supply a playtester email address; but they *do* set the map's status to 'Awaiting Playtesting'.

The map will then appear on MapHaven's front page, under 'Needs Playtesting', for a month. People can look at the map's details, and download it. When someone downloads and plays the map, they should take a few notes while they're playing the map, and then email the mapmaker with their feedback.

When the mapmaker has got enough feedback (and when at least one playtester has completed the map), they should update the map. This involves logging in to MapHaven with the same username and password as they submitted the map with, finding their map in the search system, and following the links to modify it. They should add the email address of the playtester to the map details, and switch the map status to 'Awaiting Admin Attention' or 'Awaiting More Playtesting', depending on whether the map is ready or not. They then hit the 'submit attributes and upload file' button, and upload the updated map.

"Volunteer" playtesting works like this:

Volunteer playtesters add their name to the page below this one by following the "Playtester Details" link on that page and posting a notice with their name, email address, and map preferences.

Mapmakers email one playtester at a time, asking if the playtester will test their map. The playtester should reply within a few days. If the playtester accepts, then the mapmaker can email them the map for testing. If the playtester declines, the mapmaker can email another playtester.

Conditions of playtesting:

Before submitting a map for playtesting, the mapmaker must have played their own map to completion.

The mapmaker is expected to update the map in light of the playtester's comments.

The playtester is expected to play the map to completion, or tell the mapmaker why they didn't (e.g. "I could never get past the stack of black dragons, because when I reached it in month 3, it was too big for me, and it grew faster than my army did!") (If the playtester didn't even come close to completing the map, then it'll need to be updated and sent back for more playtesting.)

Both mapmaker and playtester would do well to read the guide below, covering the whats and whys of playtesting in more detail.

DragonSister

DragonSister@heroes.mycomport.com

Playtesting Guide

So you want to be a playtester?

The amount of effort put into playtesting a map varies widely. Some people just play the map, and report back "Yeah, it was OK." Well ... for a stock-standard, very average map, that may actually be enough. It says that the map has no glaring bugs and is winnable; and that's the minimum we require maps to be tested to before they can be posted on MapHaven.

Some people spend most of their free time for a week on one map. (Well, some maps take that long to play!) One playtester I know of has sent back reports including game-dates for every moderately significant event (averaging one a game week) - plus extracted all the text from the map and proof-read it with a professional's eye - plus sent back a decent slab of commentary on possible improvements to the map, engaging in a conversation of about six email exchanges, each email about 5K long. (The initial reports were M$Word files totalling 140K.) That's the *other* extreme of playtesting. (And a complex, unusual map.)

In general playtests should be somewhere in the middle. For most maps the playtester should talk with the mapmaker a bit, and their job is mainly to suggest improvements and identify problems that need addressing.

The playtester's primary job is to report on all the *obvious* stuff. This covers:

Places you got stuck. The more complex a map is, the more likely there are to be a number of these - especially on a map with high quest components. And when it comes to working out how much a player can work out from various obscure or not so obscure hits, or thinking of all the places where hints are needed, a mapmaker can't do much themselves - they know the map inside out, after all - and this is where playtesters are most often needed.

The other thing here is that a mapmaker knows how the map is 'meant' to be played - and it is undoubtedly the way they themselves will play the map - whereas a player new to the map is quite likely to come up with different strategies.

Bugs and problems. Any place things are obviously wrong. Any place where the AI went and did stuff it probably wasn't supposed to. Any quests you couldn't complete - and you think you should have.

Major irritations. Hopefully the mapmaker should have noticed most of these in their own playtests. But if they haven't, or if something stems from a change they made after their last playthrough, you may need to tell them that "I got tired of clicking away the 'do you want to view your journal' message every day!" or "All those big ambushes with no warning meant I had to reload several times, and it wasn't fun." (Ambushes and other overly nasty surprises are things mapmakers may not expect to bother normal players. But you don't know where they are, so they may bother you.)

Anything else that you noticed as you played.

If you do that, you'll have done 'enough'. But if you want to do more, then the mapmaker will probably appreciate it; and you should think about the following:

Story.
Does the map have one? Every map should really have at least a day 1 event setting the scene, even if it only repeats the map's description about "The old king has died and now his four sons are fighting for the throne." Multiplayer maps often have nothing more than this. (Multiplayer maps should have at least this, because in HOMM3 at least, you don't normally see a map's description when setting up a match.) Whatever the map does have should be consistent. If the map description talks about rescuing a princess, then there should be something consistent with rescuing a princess in the map - a quest, or a computer player, or something. 'Rescuing a Princess' is *not* consistent with a stock-standard 4HC map. The text of a Day 1 event shouldn't clash with the map's description - or any other text you found.

If the map has story, and you've a story-telling bent yourself, you may be able to suggest improvements; a plot point to explain something that is left unexplained (like border guards between player regions or the presence of a computer player) or something like that. On matters of story, however, Mapmakers often have clear ideas that your suggestions may clash with, so don't worry if they aren't interested in that kind of suggestion. Maybe you simply suggest that something needs explaining. Again, unless it's really glaring, don't be too fussed if the mapmaker decides not to pursue those suggestions.

Scenery.
In this area, possible problems include: large empty patches; pointless patches (sometimes it's better just to pack a region with mountains and trees); large repeditive sections (mapmaker should usually break these up, throwing flowers amongst trees, lakes amongst mountains, some hills, different mountain and tree types, possibly even inaccessible buildings, creatures and resources!). It is also wise to be alert to wide variations in area available to ostensibly equal human-playable positions, for reasons of balance.

Some people like to ensure that their scenery is physically reasonable. Swamp is not generally found on top of hills. If water is, it'd better be told off as a Crater Lake or something like it (in which case, where's the volcanic terrain? Or snow?) Rivers have sources - little streams, or piles of hills, water falls and fountains. Rivers also have seas - but perhaps it just flows off the map. Snow doesn't mix with Lava - or flowers, for the most part, but it does go on large hills or at the northern and/or southern extremes of a map.

Balance.
Maps with multiple human starting postions should generally come in two types. One, balanced - where all positions have roughly equal access to resources, artifacts, power-ups, towns, quests, creature dwellings, etc etc. If the mapmaker has specified 'Multiplayer balanced' then this is *important* and you must think about it. Note that if each position is of specific alignment, then the different postions may be balanced in different ways ... for League maps this is probably not acceptable, but League maps will be tested multiple times by multiple humans to see if one position always wins or always loses.

The other kind of balancing is Position ramping. In this case the Mapmaker should say in the map's readme file that (for example) from Red the map is Easy, from Green it is Hard and from Teal it is Expert difficulty.

Ramping.
Is the map's difficulty level reasonably consistent throughout the game? If you had to wait a few weeks for enough troops to pass a particular point, and didn't have plenty to do in the meantime, there's a problem with ramping. If the map gets appreciably harder or easier at some point, there may be a problem with ramping (or you may be doing things in an order other than the one the mapmaker intended, which happens a lot, and may mean you need to tell the mapmaker to give hints or move things around. Think about the map and decide which it is.)

It is fairly common for a map to have some fights substantially harder or easier than others - and you would have to make a judgement call as to whether or not they're out of place. If the player had to fight through a garrison of 100 ogres to get to a certain point, there shouldn't be 'overwhelming ambushes' from a mere 20 ogres. If one key fight mid-map must be played 'just so' (is of expert or champion standard) when everything before and after is geared for the normal player, that's probably a problem.

Ask yourself if the map has a reasonably consistent difficulty level. Ask yourself if the map's difficulty level is anything like the difficulty level the mapmaker has suggested it is - difficulty being another of the things a mapmaker has a hard time assessing in their own map! They may need to change either the rating or the map.

Many multiplayer maps, played single-player, get markedly easier after you take out one or two opponents. This is difficult to avoid, but it can be done using 'Computer Only' events, or other tricks. See John DiFool's 'Experts' page for suggestions.

Robustness problems.
Suppose you completed a quest correctly - but what if another player dismisses the single halfling needed for a quest, or loses him in a battle? What if the player could dismiss a creature needed for a quest, without realising the creature *was* needed for a quest? Artifacts etc are less of an issue on this score. Where creatures are required for a quest, they should be either renewable (dwelling), very tough (still dangerous if there are big fights between the creature and its quest) or readily protectable by detaching (and the player is warned to look after the creature(s).) That's one type of 'robustness' problem - there are many more! A map should generally *be* robust against most different styles of play, and different happenstances - if there is only one or two ways to get to a given point, they better be obvious and/or reliably achievable.

How to be a better playtester.

There are various things that help in a playtesting endeavour. Perhaps the first of these is to be prepared to take time over it! I personally would expect to spend as long studying a map, writing about it to the mapmaker, and thinking about it and possible improvements, as I do playing the map itself. (I spend almost that much time on any map I review, too, and I usually prefer reviewing maps - so don't go sending your maps to me for playtesting! I'll hand them on to someone else!) Playtesting has a lot in common with reviewing.

The second thing is to try and get a little perspective ... something to compare the map you're playtesting to. When the very first HOMM4 maps came out, there was no way their playtesters could have played a few other amateur maps before commenting on this one - but as at August 2002 there are enough around for a novice playtester to play a few to completion before they do their first actual playtest.

If you've never playtested a map before, and you want to do it well, start by playing about three or four amateur maps with playtesting in mind. Think about each map critically and constructively. What are its strengths? (Mapmakers love to hear about these. If you want to *review* a map, too, you want to mention these to your readers.) What are its weaknesses? How might it be improved? If you've made a map yourself, you're going to be better equipped to comment on how to improve a map, but that is far from compulsory. Choose one of the maps you've played, and send some constructive feedback to the mapmaker. You have about a one in two chance of getting an appreciative email back.

Incidentally, the ability to think about something critically and constructively is a highly employable skill. And here you are practising it on something you enjoy! :-)

However - the most important thing about playtesting is not to do it perfectly, but simply to be prepared to do it! A mapmaker who isn't willing to take the time to send a map to a playtester, listen to their feedback, and amend the map as needed, probably isn't going to produce a very good map. And any playtest more detailed than "Yeah, it was OK" is better than none! :-)

Happy gaming,

DragonSister