“Slot Analytics” is a term
that is used to cover a wide
range of analytical activity.
It is helpful to divide these
analyses into two broad
categories: Slot
Performance and Slot
Comparisons. In a previous
issue (Gaming &
Leisure, Fall 2014) we discussed slot performance
analysis techniques, or the processes
where we investigate whether a slot (paytable)
is performing as expected. Is the theoretical
hold percentage of the slot machine
matching the actual hold percentage of the
game? If not, why? In this article we are going
to discuss another area of slot analytics: Slot
Comparisons.
Slot Comparisons use analytical techniques
to determine which slots or groups of slots are
“better” or “worse” for the casino. There is a
wide range of techniques that analysts use to
make these decisions, but most analysts rely
on a standard set of metrics to make these
types of comparisons.
First we should define the scope of a slot
analyst’s job. When many of us think of slot
analysis we think very specifically: “Which
game wins us more money, the Wild Wizards
or Thor’s Hammer?” But slot analytics is
actually a far larger concept. Slot analysts are
often looking for broader patterns; they are
often comparing not just slot game titles but
also other groupings of slot machines. A
group is two or more slots for which we need
to gather information concerning their relative
performance. A group could be denominations:
Which slots are more popular – the
$0.25 games or $5.00 games? A group might
be manufacturers: Which manufacturer
makes us more money – the IGT or the
William’s games? Or a group might be an
area in the casino: Which zone gets more play
– Zone 1 by the buffet or Zone 2 by the escalators?
These are the questions often asked of
our analysts, with the final goal being to optimize
the slot floor’s overall performance.
The metrics analysts typically use to measure
the performance of a slot or group of slots can
be divided into two categories: those that
measure popularity and those that measure
profitability.
Popularity metrics measure the customer’s
willingness to play a particular game (how
often a patron might select to play one game
over another). Popularity is highly correlated
with the profitability of a game; typically the
more a slot machine is played the more profitable
that slot becomes. However, popularity
metrics have an independence from profitability,
as well. Imagine a game with 100%
payback to the customer. People could be
lined up 24 hours a day to play this game, but
that game would make no money for the casino.
Furthermore, measures of profitability
(net win) are subject to a game’s volatility and
have “statistical noise” associated them. If a
game is popular, the net win statistics will
eventually even out. Therefore, it is important
that we not only examine how much a
game wins for the casino (profitability), but
also how much our customers “like” the game
(popularity). There are three common ways a
game’s popularity can be measured: Coin in,
Games Played, and Utilization.
“Coin in” is the most common way analysts
measure a game’s popularity. Coin in indicates
the total amount of dollars wagered by
the patrons over a given timeframe. The term
is a holdover from the days when, in order to
place a wager, physical coins were inserted
into the slot’s coin acceptor. Some old school
analysts sometimes refer to this as “handle.”
Others, who recognize that the idea of physical
coins is a somewhat antiquated notion, call
this measure either “bets” or “wagers.”
Outside the United States, in Macau and
South Africa for example, analysts will refer to
this measure as “turnover.” The reason for the
term “turnover” is that as a customer plays
through their initial buy-in, they will
“turnover” this buy-in several times. The
terms “coin in,” “handle,” “bets,” “wagers,”
and “turnover,” all refer to same metric; the
total dollar value that has been wagered at a
particular slot for a given time period.
Another common measure of a game’s popularity
is “Games Played.” Coin in is measured
as a dollar value and is dependent on the
base denomination of a game. It is occasionally
useful to look at a game not just in terms of
dollar value wagered on a game, but also in
terms of how many times a game has been
played over a given time frame. This measure
is sometimes referred to as “handle pulls”
which harkens back to the day when the customer
had to physically pull a handle in order
to play the game.
Another way of measuring a game’s popularity
is to look at its “utilization.” This is a
derivative of games played and can be a useful
measure in a variety of settings. The percent
utilization is calculated by taking the maximum
amount a game can be played in a given
timeframe (hour, day, week, month, etc.) and
dividing that into how many plays actually
occurred on that game during that timeframe.
For example, let us assume we have a game
where we know the maximum rate of play is
10 games per minute; in an hour the maximum
games that could be played is 600 (10
games per minute for 60 minutes). If we find
that 300 games were played on this game in
one hour, then we can say that the game was
50% utilized (300 ÷ 600). This analysis is
often done on groups of slots as well. For
example, we can extrapolate this methodology
to include areas or zones of a casino – Zone 1
was 80% utilized on Sunday whereas Zone 3
was only 10% utilized.
Several factors go into the popularity of a
game such as:
Where the game is located in the casino
– It is well known that in every casino there
are areas where customers tend to congregate,
and there are areas where customers tend to
avoid. The physical location of a slot can very
much affect its popularity.Hit frequency and payback rates – Some
games are configured to pay out many smaller
“quick pay” jackpots. These high-frequency
payouts allow customers to “replay” the same
funds several times, and this concept of
“turnover” will affect the measure of a game’s
popularity. At the same time, games that are
designed to keep a larger percentage of a customer’s
wagers will not allow customers to
“turnover” their funds as often. This will lead
to the game measuring as less popular. It is
also thought that how the game’s winning
combinations and pay out amounts affect
whether a customer will choose to play certain
game titles again affect its popularity.
Rate of play – How fast a game can be
played is often measured as “games per
minute.” This measure can also affect its popularity
metrics. Games typically average
between 6 and 8 games per minute; some
games titles can be played far faster, perhaps
in excess of 20 games per minute.
Novelty – When new game titles are
released, there is often a surge of customer
play. This popularity will usually diminish
over time and should be watched closely.
Game aesthetics and cabinet design –
Game manufactures are often updating their
game and cabinet designs, giving them modern
curves and sleeker, sometimes more
“sporty” appearances. They will also add neon
lights, use flashier graphics, and make deeper
and richer sound effects. These are all
attempts to attract patrons and increase the
popularity of the game.
The “X-factor” – Many analysts will admit
that there are some factors that affect the popularity
of a game that defy immediate understanding.
Sometimes they are forced to conclude
that we do not know why one game is
vastly more popular than another when all
known variables are similar.
Measuring and quantifying the popularity
of a slot or group of slots is a critical piece of
slot analysis, but it is only one part of the analytical
process. Equally important, if not more
important, is the profitability of slot
machines. The casino business is not just a
popularity contest; it is an enterprise that
demands profitability.
The only measure of a game’s profitability is
its win (or loss) over a specific timeframe. The
problem is that a slot machine’s “win” has
become an increasingly more elusive metric to
define. As technology develops, new factors
such as ticket-in, ticket-out, and free-play
have to be considered when calculating casino
profits (Gaming & Leisure, Fall 2014).
Net Win is defined as the net of all funds the
customer inserted into the slot cabinet such as
bills, tickets and electronic funds, less all the
funds the costumer received back such as ticket
out, hand paid jackpots, or electronic funds
uploaded into his/her account. Or simply all the
money s/he took out of his wallet or purse less all
the money the customer put back into his/her
wallet or purse. This difference is the casino win.
NET WIN = (BILLS IN + TICKET IN +
eFUNDS IN) – (TICKET OUT + HAND
PAID JP + EFUNDS OUT)
With the advent of slot systems, calculating
win using “Coin in – Coin out” became practical.
In this formula, we define “coin in” as
the total amount wagered regardless of the
source of the buy-in (bills, coin, tickets or
eFunds), and we define “coin out” as the total
amount paid out to the customer regardless of
how it was dispensed (coin to the tray, credits
to the meter, tickets issued or eFunds
uploaded and hand paid jackpots). This win
calculation is often referred to as “Machine
Win” or “Statistical Win” or sometimes as
“Actual Win.” Ideally the “Machine Win” and
“Net Win” results should be identical.
MACHINE WIN = COIN IN – TOTAL PAYOUTS
(note: total payouts includes hand pays &
coin out and occasionally customer disputes)
The Machine Win calculation has the
advantage of simplicity. Just two components:
coin in and coin out. However, the accuracy of
these numbers is totally reliant on the slot system
and game meters. Slot game meters and
the slot accounting systems are subject to a
variety of environmental challenges. Game
meters rollover, game meters are often “RAM
cleared” back to zero, new games may come
online with residual meter values still active,
slot systems may go offline, networks may
scramble data, and historical data can be lost.
It takes the constant vigilance of a dedicated
slot auditor to monitor and correct the data
when necessary.
The Net Win calculation, while more
complex, has the advantage that the components
that make up the calculation can be
independently verified through multiple
sources. Bills and tickets can be physically
counted, hand paid jackpots and cancel
credit hand pays can be physically tallied
from paperwork, game ticket meters can be
compared to the system ticket database, etc.
Where the net win data can be verified, the
machine win calculation is solely dependent
on the slot accounting system.
Analysts can choose which of these calculations
suit their specific needs but, ideally, the
net win and machine win calculations should
yield nearly identical results.
After a game or group of games’ win is
determined, there may still be adjustments to
the win. Depending on the requirements, it
may be appropriate to adjust for free-play, or
make allowances for large progressive jackpots
(see Gaming & Leisure, Summer 2014), or
reduce a game’s win by any leasing fees that
may be associated with the slot. Once the final
win numbers are determined, we can begin
the process of analyzing and comparing our
groups of slots.
There are several methodologies that analysts
can use to examine, investigate, and ultimately
determine what actions might be
taken to optimize the performance of the casino’s
slot operations. Some of the analysis
methods are:
• Direct Comparison
• Ranking and Scoring
• Relative Performance
• Percentage Contribution
Direct Comparison
Direct comparison analysis is done when two
groups of slots need to be compared so that
determinations can be made on the relative
performance of each group. Examples
include: Are Bally’s games more popular
than IGT’s games? Which games are more
profitable – progressive or non-progressive
slots? Which of two gaming titles perform
better?
Let us investigate whether the Wild
Wizards or Thor’s Hammer gaming titles
are performing better. Our slot system
reports that, for the full month of November,
our 15 Wild Wizards games won $81,250
(profitability) with $1,250,000 of coin in
(popularity). On the 8 Thor’s Hammer
games, we only won $52,000 with $860,000
of coin in. One might conclude that it lookslike the Wild Wizards are better because we
won over $29,000 more money. However, we
have to account for the fact that we have 15
Wild Wizards game and only 8 Thor’s
Hammer games. If we look at this on a winper-
unit and coin in per unit basis, the Wild
Wizards’ average per unit for the total
month are $5,416 win and $83,333 coin in,
whereas Thor’s Hammer averages $6,500
win and $107,500 coin in per unit. So on a
per unit basis, the Thor’s Hammer slots performed
better.
But wait! Before we call our slot technical
team and have them convert some Wild
Wizards into Thor’s Hammer games, let’s
look a little deeper. As it turns out, we did
not install the Wild Wizards games until
November 10th. The Thor’s Hammer games
were on for the full month. So we need to
account for this as well.
This leads us to the “gold standard” of all
performance metrics: Win per unit per day,
and coin in per unit per day. These are abbreviated
in a variety of ways: WPD and CPD
are among the most common. To calculate
WPD and CPD, we sum up how long each
game in the group has been available for play
in the period we are looking at. In our example,
we are dealing with a single month. This
will yield “unit days” or “slot days.” This is
then divided into the total win and coin in to
get WPD and CPD.
In our example:
Wild Wizards
15 slots + 20 days online = 300 unit days
Thor’s Hammer
8 slots + 30 days online = 240 unit days
So WPD and CPD are calculated as:
Wild Wizards
WPD = $81,250 / 300 = $270.83 and CPD
= $1,250,000 / 300 = $4,166.66
Thor’s Hammer
WPD = $52,000 / 240 = $216.66 and CPD
= $860,000 / 240 = $3,583.33
Our initial review of the data suggested that
the Thor’s Hammer performed better than
the Wild Wizard. However, when we
account for both the number of games AND
the number of days that the games were
operational, we find that the Wild Wizards is
both more popular and more profitable than
the Thor’s Hammer game.
Ranking and Scoring
Direct Comparison slot analysis works when
we are comparing only two groups of slots
such as two game titles or two manufacturers.
When we compare multiple groups of slots
such as all manufacturers, or all slot denominations,
or all zones/areas in a casino, a ranking
or scoring system is a better approach.
For example, let’s compare the performance
of all banks of slot machines in our casino. In
this example, we are using the term “bank” to
describe a cluster of slots that are located
throughout the casino. A bank is typically 4
to 16 slots lined up back to back, such that we
have banks that are 2 x 4 (8 Slots), 2 x 8 (16
slots), etc. A bank might be a line of slots
across a wall or at a bar, or even a group of
slots configured in a circle. It is often useful to
know how well each of these banks is performing
so that decisions on how to best optimize
the total slot performance can be made.
As noted above, the best metric for analyzing
a group of slots is to calculate the win per
unit per day (WPD) and the coin in per unit
per day (CPD). These metrics account for both
the difference in the number of slots in each
group AND the number of days each slot in
the group was active. So the performance of a
bank of 16 games that had been on the floor
for 200 days this calendar year could be directly
compared to a bank of 12 games that had
been on the slot floor for 150 days.
In a ranking system for banks, we would
simply calculate the WPD and CPD for each
bank and them sort the results. This would
then give a simple ranking of our most and
least profitable or popular banks.
To get a deeper look at the performance of
these banks we could simultaneously examine
both profitability (WPD) and popularity
(CPD). One simple approach to this would be
to first rank each bank based on WPD.
Assume we have 50 banks in our casino. We
then number each bank 1 to 50, 1 being the
highest WPD, and 50 being the lowest WPD.
We would then repeat this for CPD. By
adding these rankings together, we would get
a score that accounts equally for both profitability
and popularity (the lower the overall
score, the better the performance). Others may
choose to rate WPD higher than CPD and add
a weighting factor to emphasize WPD when
determining a performance score.
More complex scoring systems can be
derived. All of these scoring mechanisms are
designed to calculate a value that can be used
to quantify the performance of groups of slots
in relation to each other. These scoring systems
usually include components based on
both WPD (measure of profitability) and
CPD (measure of popularity).
Relative Performance
Suppose the VP of slots wanders into our analyst’s
office and says, “Back in November, I
bought 15 Wild Wizards. How are they
doing?” Well, we could tell him they are doing
better than the Thor’s Hammer games, but it
is likely he wants a better answer than that.
When we are examining a single group of slots,
be it a game title, a manufacturer’s games, or a
denomination, we need a method of determining
its relative performance. The best measure
of a game’s relative performance is to compare
it to the casino-wide WPD or CPD.
The casino win per unit per day is perhaps
the most important number in the casino’s
slot operations department, and perhaps the
most important number in the entire casino.
Slot executives dream of (or are haunted by)
this number on a daily basis. It is calculated
like any other WPD (or CPD) number, but
uses total number of unit days for ALL slots in
the casino, and the total casino win (or coin
in). Due to slot floor configuration changes,
we have to include all games – past and present
– when we derive this all-important number.
This number is often calculated using
daily, weekly, monthly and yearly totals.
Let’s assume our casino enjoyed a WPD of
$233.85 for November. So, when we want to
examine the relative performance of the Wild
Wizards, we can compare the WPD (or CPD)
of the Wild Wizards to that of the entire
casino. If the WPD of the Wild Wizards is
higher than the casino average, then the Wild
Wizards are performing well. To quantify
this, analysts often look at the group’s WPD
as a percentage of the casino WPD. This is
done by taking the difference between the
group’s WPD and the casino’s WPD, and
dividing that by the casino’s WPD.
(GROUP WPD – CASINO WPD) ÷ CASINO
WPD
A positive percentage means that the group is
performing better than the casino average,
whereas a negative percentage means the
group is performing worse than the casino
average. Let’s look at the relative performance
of the Wild Wizards. From our previous
example, we know the WPD for WildWizards was $270.83 and the casino wide
average was $233.85. The relative performance
is:
($270.83 – $233.85) ÷ 233.85 = .158 or
about 16%. So we can tell our VP that the
Wild Wizards are performing 16% higher
than the casino average.
There are many variations on this process. We
can repeat this for CPD, or we can compare
the WPD (or CPD) to other groups. Perhaps
we want to look at the relative performance of
the Wild Wizards to the averages for the
zone/area the slots are located in, or we want
to compare their relative performance to the
other games of the same denomination.
Relative performance analysis is used when we
want to compare a single group’s performance to a
broader group’s performance (casino, manufacturer,
zone, or denomination), and is a useful method for
quickly quantifying a group’s performance.
Percent Contribution
Percent contribution is an analysis method where
we compare the group’s percentage make-up of the
casino floor to its percentage contribution to either
win or coin in. If a group of slots makes up 15% of
the casino floor, then we would expect the group of
slots to make up 15% of both the total casino win
and the total casino coin in.
Let’s take two slot manufactures,
Slots R’ Us and Best
Slots USA. Our simple casino
has 100 slots; 25 slots
were purchased from Best
Slots USA and 75 slots were
purchased from Slots R’ Us.
The expectation is that 75%
of win should come from the
Slots R’ Us games and 25%
of win should come from our
Best Slots USA games.
From our slot system, we get
the following data:
• Total Casino Win for
November: $701,500
• Slots R’ Us Win for
November: $410,894 =
58.6% of win ($410,894 ÷
$701,500)
• Best Slots USA win for
November $290,606
= 41.4 % of win (290,606
÷ $701,500)
When we look at the slots’ percentage contribution
to the total casino win we see that
although Best Slots USA makes up only 25%
of the slot floor mix, they are responsible for
41.4% of the win. Maybe it’s time to call up
their sales rep and gets some more Wild
Wizards?
Percent contribution is a quick way of
revealing whether a group of slots is contributing
“their fair share” to the net win and
coin in against the entire casino. It can be used
to quickly identify under-performing and
over-performing groups.
****
Slot Comparisons help us to make decisions
on how to best optimize the slot floor’s overall
performance. Groups of slots could be
zones, denominations, manufacturers, game
titles or any number of other combinations.
Analysts use metrics that both measure a
game’s popularity like coin in, games-played,
and utilization, as well as metrics that measure
a game’s profitability like net win and
machine win. The best metrics of popularity
and profitability are win per unit per day and
coin in per unit per day. These are the best
metrics because they account for the differences
in the number of slots in each group
and the amount time each slot has
been operational. Four popular methods for
looking at this data are:
Direct Comparison: To investigate and compare
ONLY two groups of slots
Ranking and Scoring: To rank multiple
groups of slots and quantifiably score them
Relative Performance: To compare a single
group of slots to a broader groups of slots
Percent Contribution: To compare the
expected contribution to win or coin in based
on the percentage make-up
These analysis methods, while not the only techniques
available, give casino operators quantitative
data to evaluate both the popularity and
profitability of their casino slot floor.
Josh Cantrell is Founder of and Senior Slot Systems
Consultant for Sierra Gaming Consultants
(www.sierragamingconsultants.com). He specializes
in slot operations, slot accounting / analysis,
and slot system support. He has over 25
years of experience working in casino markets
around the world including those in Canada,
Mexico, Macau, South Africa, and throughout
South America.