Vistula Encounter Tables Analysis - the Character of a River

Apart from the random tables in Death on the Reik, the only usable riverine random encounter tables I've come across so far are those from the T2k adventure Pirates of the Vistula. Let's then have a look at the "General" Encounter tables for each of the 4 main sections of the stretch of the Vistula covered by the module to see what can we extrapolate:

Pirates of the Vistula General Encounter Tables 

Note: there are more specific tables for minor stretches of the river in the module, usually associated with the riverine towns or particular geographical areas, but I'm not going to analyse that here. Likewise, I'm not looking at the "Shore" section of the tables as the module notes these are intended for use in the hinterland sections when the party is exploring ashore, not for the riparian sections.     

Looking at these section River General Encounter tables, based on the distribution of a 2D6 roll, there are a few common features and associated probabilities:

• 2 is always a Hazard 2.78%
• 5-7 is always No Encounter 41.67%
• 9-12 are always Hazards 27.78%

This accounts for ~73% of the "encounters", although effectively ~40% (~5/12 or ~4/10) are actually no encounter. Hazards as a group, however, are encountered at least 30% (4/12 or 3/10) of the time, which seems quite high on first thought, although I guess the Vistula could be considered "remote" and "unimproved" - if anything it could be considered "hazardous" (worse than "unimproved") due to the numerous ruined structures and other post-apocalyptic debris (more on different waterways in a later post, however). 

Two outcomes show some minor variability, but I'm unsure if this was intentional or not.

• 3 is usually Hazard except for Section 2 (No Encounter) 5.56%
• 4 is No Encounter except Section 4 (Mud shoal) 8.33%

Both of these outcomes are relatively rare on a 2D6 distribution.

Only one of the outcomes has any real variability across the tables, resulting in "No Encounter" for the first and third sections of the river, alternating for the other 2 sections.

• 8 is variable across sections (No encounter, Mud shoal or Boat) 13.89% 

It's this particular outcome that influences the tables given the high frequency on a 2D6 roll. This is effectively the "discretionary" outcome of the table, used to increase the probability of one of the three outcomes by about 10%, either as No Encounter, Hazard or Encounter.

Three Examples 

From herein the maths is not exact and I'm rounding to 5% (1 in 20) intentionally here for ease.

Let's assume for a first example that the 8 result outcome is an Encounter for the set of calculations below. So the base table without the variability then roughly leaves the base chance of outcome as:

• No Encounter     ~50%
• Hazard                ~35%
• Encounter           ~15%  

Using these estimates, a party travelling down the Vistula will note that hazards are about twice as common as actual encounters with humans. As per the module, checks are done for the morning and the evening when travelling (unlike DotR, there is no default option for night travel or a separate table), which results in the following combinations for each day:

• A Hazard 36%
• Uneventful 25%
• An Encounter 16%
• 2 Hazards 12%
• An Encounter & a Hazard 10%
• 2 Encounters 2%

This means there's about a ~60% chance (6 in 10) each day of at least one Hazard, a 28% chance (3 in 10) of at least one Encounter, and a 25% chance (1 in 4) of an Uneventful journey each day.

In a second example, let's assume the 8 result is a "No Encounter" ie. a less populated region.

So the base table without the variability then roughly leaves the base chance of outcome as:

• No Encounter     ~60%
• Hazard                ~35%
• Encounter             ~5%  

With these estimates, a party travelling down the Vistula will note that hazards are much more common than actual encounters with humans. As per the module, checks are done for the morning and the evening when travelling (unlike DotR, there is no default option for night travel or a separate table), which results in the following combinations for each day:

• A Hazard 42%
• Uneventful 36%
• An Encounter 10%
• 2 Hazards 12%
• An Encounter & a Hazard 4%
• 2 Encounters 0%

This means there's about a ~60% chance (6 in 10) each day of at least one Hazard, a 14% chance (1 in 10) of at least one Encounter, and a 36% chance (4 in 10) of an Uneventful journey each day.

In a third example, let's assume the 8 result is a "Hazard" ie. a treacherous stretch.

So the base table without the variability then roughly leaves the base chance of outcome as:

• No Encounter     ~50%
• Hazard                ~45%
• Encounter             ~5%  

In this example party travelling down the Vistula will note that hazards are even more common than actual encounters with humans. As per the module, checks are done for the morning and the evening when travelling (unlike DotR, there is no default option for night travel or a separate table), which results in the following combinations for each day:

• A Hazard 46%
• Uneventful 25%
• An Encounter 4%
• 2 Hazards 20%
• An Encounter & a Hazard 6%
• 2 Encounters 0%

This means there's about a ~70% chance (7 in 10) each day of at least one Hazard, a 10% chance (1 in 10) of at least one Encounter, and a 25% chance (2 in 10) of an Uneventful journey each day.

An Alternate Distribution Method

Let's look at using an alternative "d4+d8 roll" distribution on the same table to generate a flatter distribution for the first example above (a result of 8 is an Encounter and reflects a 12.5% probability):

• No Encounter     ~40%
• Hazard                ~40%
• Encounter           ~20%  

This is calculated using the "Anydice" website and its probabilities mapped from above.

This seems a lot neater somehow and is similar enough to the standard method - again Hazards are about twice as common as encounters with humans but equal to No Encounter and this results in the following distribution of probabilities:

• A single Hazard 32%
• 2 Hazards 16%
• An Encounter & a Hazard 16%
• An Encounter 16%
• Uneventful 16%
• 2 Encounters 4%

This second distribution then results in a 64% chance each day of at least one Hazard (4 in 6),  a 36% chance of at least one Encounter (2 in 6) and a 16% chance (1 in 6) of an uneventful journey for that day. The overlap of a Hazard and an Encounter is also neatly 16% (1 in 6). Even with the adjusted probabilities, a cruise down the Vistula is going to be quite eventful and full of hazards.

The other two examples result in an Encounter chance 10% and either No Encounter or Hazard of either 40% or 50% depending on what the result of the "8 outcome" is chosen to be. I won't work these out in full here, but they similarly give options for either a more desolate or treacherous stretch of river respectively.

Note: Interestingly, the 2 in 6 chance correlates with the chance of a random encounter (hazard or creature) in an OSR style game such as Labyrinth Lord for rivers (2 in d6), although the original rules suggest checking against this chance 2-3 times per day. This actually generates a much higher 46% chance (3 in 6) or 31% chance (2 in 6) of "No Encounter" each day (based on 2 and 3 rolls respectively). The split between Hazards and creature encounters isn't specified in older games, although traditionally early encounter tables were exclusively creature based.

Comparisons with the Reik

Using either distribution, these three examples provide significant variation from the base layout of the table but by extrapolation from the module as written, regardless of which stretch of the river is being travelled, a cruise down the Vistula is still going to require a lot of random obstacles dodging or debris clearing, with only infrequent stretches of an uninterrupted waterway and some occasional human encounters (river-borne or on the shore).

All this is without the set-piece encounters at narrows, townships and bridges (intact or ruined).

Compared to the stretch of the River Reik presented in Death on the Reik, with a calculated 30-40% encounter rate per check during the day (refer to the "River Life of the Empire" booklet), the Vistula is therefore much, much less busy but with a lot more potential hazards (the Reik has only a <10% calculated Hazard rate). This makes sense as the Reik is densely populated, highly patrolled and significantly improved and maintained without many natural hazards, whereas the post-WW3 version of the Vistula portrayed in the module is only sparsely settled, virtually unpatrolled (except for the Korsarz near Warsaw) and not only littered with ruined structures (bridges and docks) but seems intrinsically treacherous by nature (mud shoals, sandbars and rocks). 

The two rivers have vastly different characters in terms of population, control and improvements and this is therefore appropriately reflected by their random encounter tables. They are not quite at the extremes but provide a good contrast at different ends of the spectrums of character.

I think this helps explain what makes each river feel different in terms of "personality".

Concluding Thoughts

So what then makes up a river's character?

Let's see if we can pick out some common themes from the analysis above. A river can be...

1. Navigable vs Unavigable
2. Populated vs Remote
3. Patrolled vs Unpatrolled
4. Improved vs Unimproved
5. Tame vs Hazardous

These dimensions should allow me to model a few key characteristics of rivers (as waterways ie navigable) that will help construct different riverine encounter tables going forward. 

For example, the upstream Utdoo section of the river winding through the Amedio in UK6 is:

1. Unavigable
2. Remote
3. Unpatrolled
4. Unimproved
5. Tame

However, the Black River from Corvis to Merywyn in the Iron Kingdoms is likely:

1. Navigable
2. Neither Populated nor Remote
3. Patrolled
4. Unimproved
5. Neither Tame nor Hazardous

It's a start in any case and something I'll pick up on in a later post...